Growth of InAs Quantum Dots on GaAs Nanowires by Metal Organic

Aug 17, 2011 - InAs quantum dots (QDs) are grown epitaxially on Au-catalyst-grown GaAs nanowires (NWs) by metal organic chemical vapor deposition (MOC...
0 downloads 11 Views 3MB Size
LETTER pubs.acs.org/NanoLett

Growth of InAs Quantum Dots on GaAs Nanowires by Metal Organic Chemical Vapor Deposition Xin Yan,† Xia Zhang,*,† Xiaomin Ren,*,† Hui Huang,†,‡ Jingwei Guo,† Xin Guo,† Minjia Liu,† Qi Wang,† Shiwei Cai,† and Yongqing Huang† †

State Key Laboratory of Information Photonics and Optical Communications, Beijing University of Posts and Telecommunications, Beijing 100876, China ‡ School of Electronic Science and Technology, Dalian University of Technology, Dalian 116024, China ABSTRACT: InAs quantum dots (QDs) are grown epitaxially on Au-catalyst-grown GaAs nanowires (NWs) by metal organic chemical vapor deposition (MOCVD). These QDs are about 10 30 nm in diameter and several nanometers high, formed on the {112} side facets of the GaAs NWs. The QDs are very dense at the base of the NW and gradually sparser toward the top until disappearing at a distance of about 2 μm from the base. It can be concluded that these QDs are formed by adatom diffusion from the substrate as well as the sidewalls of the NWs. The critical diameter of the GaAs NW that is enough to form InAs QDs is between 120 and 160 nm according to incomplete statistics. We also find that these QDs exhibit zinc blende (ZB) structure that is consistent with that of the GaAs NW and their edges are faceted along particular surfaces. This hybrid structure may pave the way for the development of future nanowire-based optoelectronic devices. KEYWORDS: Nanowire, quantum dots, MOCVD, adatom diffusion, critical diameter, zinc blende

S

emiconductor NWs have attracted enormous attention in recent years due to their great potential as building blocks for future devices such as nanolasers,1,2 field-effect transistors,3 LEDs,4 solar cells,5 and data-storage devices.6 As an important element of high-performance nanoscale devices, NW heterostructure has caused more and more interest. Axial NW heterostructures as well as core shell and core multishell heterostructures have been widely investigated.2,10 13 Recently, an attractive structure has emerged that combines NWs with QDs. A highly efficient single-photon source as well as a single quantum dot NW LED has been fabricated successfully by inserting a QD in a NW,7,8 and a solar cell with QDs blocking in the space among NWs also raises the energy conversion efficiency to approximately 1%.9 Both of the two NW-QD structures show great potential in fabricating multifunctional and high-performance devices. It is well-known that in the planar epitaxy, if the lattice constant of the film is much larger than that of the planar substrate, the growth may follow a Stranski Krastanow (S-K) mechanism and QDs are formed.14 As the lateral surface of a NW is big, it is possible that QDs are formed on the side walls. Actually, formations of Ge QDs on Si NWs and MnAs QDs on InAs NWs have already been reported.15,16 Recently, formation of InAs QDs on GaAs NWs was also realized by molecular beam epitaxy (MBE) using galliumassisted catalyst-free method.17 In their experiment, InAs QDs were formed on {110} side facets of GaAs NWs by previously depositing an AlAs shell layer, which overcame the obstacle that InAs QDs do not form on GaAs {110}-oriented surfaces.18 In this work, self-assembled InAs QDs were grown epitaxially on {112} side walls of GaAs NWs by MOCVD via Au-catalyst vapor liquid solid (VLS) mechanism. The distribution, morphological r 2011 American Chemical Society

and structural characteristics of the QDs were discussed in detail. The critical diameter of the GaAs NW that is enough to form InAs QDs was estimated experimentally. We also developed a growth mechanism to explain the formation of QDs on NWs. The Au-assisted VLS growth was performed by using a Thomas Swan CCS-MOCVD system at a pressure of 100 Torr. Trimethylgallium (TMGa), trimethylindium (TMIn), and arsine were used as precursors. The carrier gas was hydrogen. Growth was carried out as the following steps. (1) An Au film with a thickness of 4 nm was deposited on GaAs (111)B substrate by magnetron sputtering; (2) the Au-coated GaAs substrate was loaded into the MOCVD reactor and annealed at 650 °C under arsine ambient for 300 s to form the Au Ga alloyed particles as catalyst; (3) after ramped down to the desired temperature, GaAs NWs were grown at 440 °C for 400 s. TMGa was switched off after growth while the flow rate of AsH3 was kept constant at 2.87  10 3 mol min 1; (4) after raising the temperature to 475 °C, growth of InAs QDs was initiated for 30 s with a TMIn flow rate of 11.3 μmol min 1 and a AsH3 flow rate of 71.8 μmol min 1. Morphologies of the samples were characterized by field emission scanning electron microscopy (FESEM) and the structural characteristics of the samples were investigated by transmission electron microscopy (TEM), respectively. TEM specimens were prepared by ultrasonicating the samples in ethanol for 5 min, followed by spreading drops from the suspension onto a holey carbon/Cu grid. Received: June 28, 2011 Revised: August 8, 2011 Published: August 17, 2011 3941

dx.doi.org/10.1021/nl202190n | Nano Lett. 2011, 11, 3941–3945

Nano Letters

Figure 1. (a) SEM image of the NWs with a 20° view from surface normal direction; (b) cross sectional view SEM image of the NWs; (c) SEM image of a single NW from the top.

SEM images of the NWs are shown in Figure 1. From Figure 1a,b, we can see that most of the NWs are vertical to the substrate except some thin and bent ones like whiskers. As seen in Figure 1a, diameters of the NWs are quite different rather than uniform, which is due to a broad size distribution of Au Ga particles as a result of random agglomeration of Au atom.19 As estimated from Figure 1b, the diameters range from several dozens of nanometers to several hundreds and the heights are about 5 6 μm. We also find that the top parts of the NWs are straight and capped with Au particles, which indicates that InAs does not grow epitaxially on top of the GaAs NW as the Au particle usually moves sideways that commonly observed in the growth of axial GaAs/InAs NW heterostructure.11,20 Thus we think that InAs may attach to the side walls, and this is supported by Figure 1c. From the top view image, we can clearly see that the surface of the NW sidewalls is rather rough, decorated with a lot of small bumps. Using the {110} cleavage plane of the substrate

LETTER

as a reference, we can determine that the NW has {112} type side facets, which commonly appears in the case of Au-catalyzed NWs.21 Figure 2a shows an enlarged SEM image of the lower parts of the NWs. We find a lot of small dots assembling at the bases and arranging along the side walls for all of the four wires with different diameters (the diameters are 720, 635, 315, and 170 nm, respectively, from left to right). To take a close watch on the small dots, we choose a single NW with a diameter of 635 nm and observe it from the bottom up carefully. Figure 2b e shows the areas of b, c, d, and e, respectively, as indicated in Figure 2a. From Figure 2b, we can see that the dots are rather dense and even join up with each other at the base with diameters ranging from 10 to 30 nm. At the upper part near the root, the dots are aligned in several ribbons along the side walls in the growth direction, as shown in Figure 2c. The area of d is close to 2 μm from the substrate. From Figure 2d, we can only find several dots on the surface, sparse and small. As we move to the area of e, which exceeds 2 μm from the substrate, no dots exist and the lateral surface is rather smooth as shown in Figure 2e. Figure 2f shows the SEM image of a single NW with a much smaller diameter of about 170 nm, and we also find several dots on the side walls, although much sparser than the above-mentioned one. From the SEM images of Figure 2 and TEM images of Figure 3, diameters of the dots are measured to be about several dozens of nanometers while the heights are several nanometers, and both of the dimensions are similar to that of the InAs QDs formed on the GaAs planar substrate.22 As the size of the QD is rather tiny, it is very difficult to determine the composition by Energy-dispersive X-ray Spectroscopy (EDS). But we can draw a conclusion that these dots are InAs QDs as no dots were found in our previous GaAs NWs grown under the same conditions just without step 4 in the above-mentioned experimental details.19 It is well-known that for VLS growth, there are two major contributions, that is, direct impingement of the precursors onto the alloy droplet and adatom diffusion from the sidewalls and substrate surface to the top.23 25 It has been demonstrated that the adatom diffusion will result in lateral overgrowth and tapering when the wire length is longer than the diffusion.26,27 From Figure 1b, we can see that the GaAs NWs are highly uniform in diameter. This indicates that the diffusion of Ga adatoms makes little contribution, which is consistent with our preceding results.12,28 Considering that the growth temperature of InAs QDs is raised up to 475 °C, the adatom diffusion mechanism is probably enhanced while the direct impingement drops off.29,30 Furthermore, it is well-known that In species have a larger diffusion length than Ga species. Kim et al. has demonstrated that the Ga diffusion length is less than 2 μm on the (111)B GaAs surface while the In diffusion length is about 6 μm.31 Thus, In adatoms are more likely to diffuse from the substrate surface and deposit on the sidewalls, resulting in QDs attaching to the surface. As the lateral surface of the NW is big and the lattice mismatch between GaAs and InAs is rather large (about 6.7%), the growth probably tends to follow an S-K mode and InAs QDs are formed, which is similar to the growth of InAs QDs on the planar GaAs substrate.14,22 It should be noted that the volume per unit surface area of a NW is quite smaller than that of a standard planar substrate. Thus the NW could share more strain energy caused by the lattice mismatch.32 In addition, the nanosized curve surface of the NW could also help to release the lattice strain.33 The resulted reduction of the strain stored in the epitaxial layer would lead to a larger critical thickness for the QD formation. 3942

dx.doi.org/10.1021/nl202190n |Nano Lett. 2011, 11, 3941–3945

Nano Letters

LETTER

Figure 2. (a) An enlarged cross sectional view SEM image of the lower parts of the NWs; (b e) SEM images of different parts of a single NW with a diameter of 635 nm, corresponding to the areas of b, c, d and e as indicated in a, respectively; (f) SEM image of a part of a single NW with a diameter of 170 nm, corresponding to the area of f in (a). Dots are indicated by arrows.

Figure 3. (a)∼(f) TEM images of a single NW with diameters of 160 nm, 280 nm, 120 nm, 113 nm, 110 nm, and 78 nm, respectively. QDs in (a) are indicated by black arrows.

Therefore, we speculate that the critical thickness for the formation of InAs QDs on GaAs NWs would be thicker than that on the standard GaAs {100} substrate. The latter was estimated to be 1 2 ML as previously reported.34 The formation of QDs on the surface of NWs has also been observed in the growth of germanium (Ge) on silicon (Si) NWs and MnAs on InAs NWs.15,16 As reported by Uccelli et al., InAs QDs were formed on {110} side facets of GaAs NWs by previously depositing an AlAs shell layer.17 While in our experiment, the process is simplified to some extent as no AlAs shell layer was deposited, and InAs QDs were formed on {112} type side facets rather than {110}. The great difference probably attributes to the different growth methods (catalyst-free MBE and Au-catalyst MOCVD) as well as the different properties of different side facets ({110} and {112}), although which still needs further investigations. As mentioned before, the QDs become sparser and smaller toward the top of the NWs, and finally disappear at a distance of about 2 μm from the base. This could be attributed to two

mechanisms as follows: on one hand, as mentioned above, the InAs QDs are formed by adatom diffusion from the sidewalls and substrate surface. The diffusion length along the NW is surely limited, while the number of In adatoms may dramatically decrease on the way to the top as the GaAs NW could effectively consume and deplete In adatoms.35 On the other hand, the Au particle may act as a collector of the diffusing In adatoms and some of the adatoms on the sidewall, especially those near the Au particle, are probably collected.36 We have investigated the height of the QD coverage (or the distance from the substrate to where the QDs starts to disappear along the NW length) of each individual NW among several selected ones with different heights. For the four NWs with heights of 4.8, 4.9, 6.0, and 6.5 μm, the heights of the QD coverages are 2.0, 2.2, 2.2, and 2.2 μm, respectively. So, it could be concluded that the height of the QD coverage seems nearly independent of the entire height of a NW. Thus, if we shorten the average height of the NWs to a smaller value around 2 μm, probably most of the NWs would be the fully QD-covered 3943

dx.doi.org/10.1021/nl202190n |Nano Lett. 2011, 11, 3941–3945

Nano Letters

LETTER

Figure 4. Distribution of the diameters that could or could not form QDs. Arrows indicate the diameters of some samples in our experiment.

ones. Also, this hybrid structure has promising applications in high-performance nanowire lasers and single-photon sources, as well as high efficiency solar cells. It is important to point out that we did not observe an axial InAs section at the top of the NW although some In atoms may be incorporated into the Au particle. Its absence is mainly attributed to the As-lacked environment with a rather low V/III ratio of about 0.23 during the growth. Under this condition, InAs could not precipitate due to the extremely low content of As soluted in the Au In alloy.37 Figure 3 shows the TEM images of several single NWs with different diameters, and those NWs can be classified into two types according to their morphologies: one is with QDs on the surface while the other without. NWs in Figure 3a,b belong to the former type with diameters of 160 and 280 nm, respectively. From Figure 3a, we can see an array of QDs along the wire, and the last QD is about 2 μm from the base. NWs in Figure 3c, d, e, and f belong to the latter type with diameters of 120, 113, 110, and 78, respectively. No QDs could be found on those NWs and their lateral surfaces are rather smooth. According to research on thin film growth, as the substrate dimension is reduced, the thickness of the strained epitaxial film will increase since the strain energy is distributed more evenly between the epilayer and the substrate.38 In this condition, the NW serves as the substrate thus the one with a larger diameter is much easier to form QDs than the smaller one. Xinlei et al. has studied the epitaxy of Ge on Si NWs theoretically and found that only when the diameter of Si wire exceeds a critical value (about 70 nm), Ge QDs could form on the surface.33 According to our statistics, the diameters of NWs with QDs are {720, 635, 315, 280, 170, and 160 nm} and the diameters of NWs without QDs are {120, 113, 110, and 78 nm}, which are shown in Figure 4. Thus we can determine that the critical diameter of GaAs NW is between 120 and 160 nm. Figure 5a shows the HRTEM image of a single QD. The diameter of the QD is about 24 nm and the height is 6 nm. The QD has an irregularly truncated pyramidal shape, the cross section of which is an asymmetric trapezium with one waist near the base much longer than the other near the top. We have observed all the QDs of the samples and find they have nearly the same shape. We attribute this unusual phenomenon to the diffusion mechanism which would not occur in the planar epitaxy. We also find that the edges of the QD are faceted along particular surfaces. Using the [111] growth direction as a reference, three facets can be determined as (111), (112), and (110). Figure 5b shows the

Figure 5. HRTEM images of the QDs. (a) HRTEM image of a QD with the lines illustrating its crystal planes; (b) HRTEM image of two adjacent QDs with red arrows indicating the twin and stacking faults.

HRTEM image of two adjacent QDs. There exists a {111} type twin in the left QD, while the right one contains several (111) stacking faults. These stacking faults are likely bounded by Shockley partial dislocations at the core/island interface that act as misfit dislocations.15 The InAs QDs exhibit ZB structure which is the same as that of the GaAs NW, and this is consistent with the structure that a GaAs wire sheathed with an InAs shell.39 In summary, self-assembled InAs QDs were grown epitaxially on the surface of GaAs NWs by MOCVD. The diameters of the QDs range from 10 30 nm while the heights are several nanometers. The QDs are very dense at the base of the NW and gradually sparser toward the top until disappearing at a distance of about 2 μm from the base. We conclude that the QDs are formed by adatom diffusion from the substrate and the sidewalls, and the GaAs NW as well as the Au particle has a strong effect on the diffusion 3944

dx.doi.org/10.1021/nl202190n |Nano Lett. 2011, 11, 3941–3945

Nano Letters length. The critical diameter of the GaAs NW is between 120 and 160 nm. The crystal facets of the QDs are determined from the HRTEM images and the twins and stacking faults in the QDs are also observed. This hybrid structure shows great promise in future nanowire-based optoelectronic devices such as high-performance lasers and single-photon sources, as well as high efficiency solar cells.

’ AUTHOR INFORMATION Corresponding Author

*E-mail: (X.Z.) [email protected]; (X.R.) [email protected].

’ ACKNOWLEDGMENT This research was supported by the National Basic Research Program of China (2010CB327600), the National Natural Science Foundation of China (90201035 and 61077049), New Century Excellent Talents in University (NCET-08-0736), the National High Technology R&D Program of China (2009AA03Z417), the Fundamental Research Funds for the Central Universities (2011RC0401) and the 111 Program of China (B07005). ’ REFERENCES (1) Huang, M. H.; Mao, S.; Feick, H.; Yan, H.; Wu, Y. Y.; Kind, H.; Weber, E.; Russo, R.; Yang, P. D. Science 2001, 292, 1897. (2) Qian, F.; Li, Y.; Gradecak, S.; Park, H. G.; Dong, Y. J.; Ding, Y.; Wang, Z. L.; Lieber, C. M. Nat. Mater. 2008, 7, 701. (3) Xiang, J.; Lu, W.; Hu, Y. J.; Wu, Y.; Yan, H.; Lieber, C. M. Nature 2006, 441, 489. (4) Minot, E. D.; Kelkensberg, F.; Kouwen, M. V.; Dam, J. A. V.; Kouwenhoven, L. P.; Zwiller, V.; Borgstr€om, M. T.; Wunnicke, O.; Verheijen, M. A.; Bakkers, E. P. A. M. Nano Lett. 2007, 7, 367. (5) Kempa, T. J.; Tian, B. Z.; Kim, D. R.; Hu, J. S.; Zheng, X. L.; Lieber, C. M. Nano Lett. 2008, 8, 3456. (6) Jung, Y.; Lee, S. -H.; Jennings, A. T.; Agarwal, R. Nano Lett. 2008, 8, 2056. (7) Claudon, J.; Bleuse, J.; Malik, N. S.; Bazin, M.; Jaffrennou, P.; Gregersen, N.; Sauvan, C.; Lalanne, P.; Gerard, J. M. Nat. Photonics 2010, 4, 174. (8) Minot, E. D.; Kelkensberg, F.; van Kouwen, M.; van Dam, J. A.; Kouwenhoven, L. P.; Zwiller, V.; Borgstr€om, M. T.; Wunnicke, O.; Verheijen, M. A.; Bakkers, E. P. A. M. Nano Lett. 2007, 7, 367. (9) Nadarajah, A.; Word, R. C.; Sant, K. V.; K€onenkamp, R. Phys. Status Solidi B 2008, 245, 1834. (10) Bj€ork, M. T.; Ohlsson, B. J.; Sass, T.; Persson, A. I.; Thelander, C.; Magnusson, M. H.; Deppert, K.; Wallenberg, L. R.; Samuelson, L. Appl. Phys. Lett. 2002, 80, 1058. (11) Paladugu, M.; Zou, J.; Guo, Y. N.; Auchterlonie, G. J.; Joyce, H. J.; Gao, Q.; Tan, H. H.; Jagadish, C.; Kim, Y. Small 2007, 3, 1873. (12) Huang, H.; Ren, X. M.; Ye, X.; Guo, J. W.; Wang, Q.; Zhang, X.; Cai, S. W.; Huang, Y. Q. Nanotechnology 2010, 21, 475602. (13) Tomiok, K.; Motohisa, J.; Hara, S.; Hiruma, K.; Fukui, T. Nano Lett. 2010, 10, 1639. (14) Heinrichsdorff, F.; Krost, A.; Kirstaedter, N.; Mao, M. H.; Grundmann, M.; Bimberg, D.; Kosogov, A. O.; Werner, P. Jpn. J. Appl. Phys. 1997, 36, 4129. (15) Pan, L.; Lew, K. -K.; Redwing, J. M.; Dickey, E. C. Nano Lett. 2005, 5, 1081. (16) Ramlan, D. G.; May, S. J.; Zheng, J. G.; Allen, J. E.; Wessels, B. W.; Lauhon, L. J. Nano Lett. 2006, 6, 50. (17) Uccelli, E.; Arbiol, J.; Morante, J. R.; Morral, A. F. ACS Nano 2010, 4, 5985. (18) Belk, J. G.; Pashley, D. W.; McConville, C. F.; Sudijono, J. L.; Joyce, B. A.; Jones, T. S. Phys. Rev. B 1997, 56, 10289.

LETTER

(19) Ye, X.; Huang, H.; Guo, J. W.; Ren, X. M.; Huang, Y. Q.; Wang, Q. J. Semicond. 2010, 31, 073001. (20) Paladugu, M.; Zou, J.; Auchterlonie, G. J.; Guo, Y. N.; Kim, Y.; Joyce, H. J.; Gao, Q.; Tan, H. H.; Jagadish, C. Appl. Phys. Lett. 2007, 91, 133115. (21) Wacaser, B. A.; Deppert, K.; Karlsson, L. S.; Samuelson, L.; Seifert, W. J. Cryst. Growth 2006, 287, 504. (22) Hsieh, T. -P.; Chiu, P. -C.; Chyi, J. -I.; Yeh, N. -T.; Ho, W. J.; Chang, W. H.; Hsu, T. -M. Appl. Phys. Lett. 2005, 87, 151903. (23) Wagner, R. S.; Ellis, W. C. Appl. Phys. Lett. 1964, 4, 89. (24) Lu, W.; Lieber, C. M. J. Phys. D: Appl. Phys. 2006, 39, R387. (25) Dubrovskii, V. G.; Sibirev, N. V.; Cirlin, G. E.; Soshnikov, I. P.; Chen, W. H.; Larde, R.; Cadel, E.; Pareige, P.; Xu, T.; Grandidier, B.; Nys, J.-P.; Stievenard, D.; Moewe, M.; Chuang, L. C.; Chang-Hasnain, C. Phys. Rev. B 2009, 79, 205316. (26) Plantea, M. C.; LaPierre, R. R. J. Cryst. Growth 2008, 310, 356. (27) Dubrovskii, V. G.; Sibirev, N. V.; Cirlin, G. E.; Tchernycheva, M.; Harmand, J. C.; Ustinov, V. M. Phys. Rev. E 2008, 77, 031606. (28) Huang, H.; Ren, X. M.; Ye, X.; Guo, J. W.; Wang, Q.; Yang, Y. S.; Cai, S. W.; Huang, Y. Q. Nano Lett. 2010, 10, 64. (29) Borgstr€om, M.; Deppert, K.; Samuelson, L.; Seifert, W. J. Cryst. Growth 2004, 260, 18. (30) Soci, C.; Bao, X. Y.; Aplin, D. P. R.; Wang, D. Nano Lett. 2008, 8, 4275. (31) Kim, Y.; Joyce, H. J.; Gao, Q.; Tan, H. H.; Jagadish, C.; Paladugu, M.; Zou, J.; Suvorova, A. A. Nano Lett. 2006, 6, 599. (32) Raychaudhuri, S.; Yu, E. T. J. Appl. Phys. 2006, 99, 114308. (33) Li, X. L.; Yang, G. W. J. Phys. Chem. C 2009, 113, 12402. (34) García, J. M.; Silveira, J. P.; Briones, F. Appl. Phys. Lett. 2000, 77, 409. (35) Paladugu, M.; Zou, J.; Guo, Y.; Zhang, X.; Joyce, H.; Gao, Q.; Tan, H.; Jagadish, C.; Kim, Y. Appl. Phys. Lett. 2008, 93, 201908. (36) Seifert, W.; Borgstr€om, M.; Deppert, K.; Dick, K. A.; Johansson,  J.; Larsson, M. W.; Martensson, T.; Sk€old, N.; Svensson, C. P. T.; Wacaser, B. A.; Wallenberg, L. R.; Samuelson, L. J. Cryst. Growth 2004, 272, 211. (37) Dick, K. A.; Deppert, K.; Samuelson, L.; Seifert, W. J. Cryst. Growth 2006, 297, 326. (38) Lo, Y. H. Appl. Phys. Lett. 1991, 58, 2311. (39) Paladugu, M.; Zou, J.; Guo, Y. N.; Zhang, X.; Joyce, H. J.; Gao, Q.; Tan, H. H.; Jagadish, C.; Kim, Y. Nanoscale Res. Lett. 2009, 4, 846.

3945

dx.doi.org/10.1021/nl202190n |Nano Lett. 2011, 11, 3941–3945