• 沒有找到結果。

Unusual optical properties of type-II InAs/GaAs0.7Sb0.3 quantum dots by photoluminescence studies

N/A
N/A
Protected

Academic year: 2021

Share "Unusual optical properties of type-II InAs/GaAs0.7Sb0.3 quantum dots by photoluminescence studies"

Copied!
4
0
0

加載中.... (立即查看全文)

全文

(1)

Unusual optical properties of type-II InAs/ GaAs

0.7

Sb

0.3

quantum dots

by photoluminescence studies

T. T. Chen, C. L. Cheng, and Y. F. Chen*

Department of Physics, National Taiwan University, Taipei 106, Taiwan

F. Y. Chang and H. H. Lin

Department of Electrical Engineering, National Taiwan University, Taipei 106, Taiwan

C.-T. Wu

Department of Material Science and Engineering, National Taiwan University, Taipei 106, Taiwan

C.-H. Chen

Center for Condensed Matter Sciences, National Taiwan University, Taipei 106, Taiwan

共Received 31 July 2006; revised manuscript received 8 November 2006; published 17 January 2007兲

The optical properties of type-II InAs/ GaAs0.7Sb0.3quantum dots共QDs兲 were investigated by photolumi-nescence共PL兲. It is found that the peak position of PL spectra exhibits a significant blueshift under a moderate excitation level. The observed blueshift can be well explained by the band bending effect due to the spatially separated photoexcited carriers in a type-II band alignment. We also found that the PL spectra exhibit a strong in-plane polarization with a polarization degree up to 24%. The observed optical anisotropy is attributed to the inherent property of the orientation of chemical bonds at InAs/ GaAs0.7Sb0.3heterointerfaces.

DOI:10.1103/PhysRevB.75.033310 PACS number共s兲: 78.67.Hc

InAs-based quantum dots共QDs兲 provide potential appli-cations in long wavelength optical-fiber communiappli-cations. To obtain the required wavelength of 1.3␮m and 1.5␮m in QDs lasers, several methods have been used to reduce the influence of compressive strain effect, such as the strain re-ducing layer共SRLs兲, stacking QDs, and surface QDs.1–3 Al-ternatively, type-II band alignment also provides an excellent way to lengthen emission wavelength.4–7For InAs QDs em-bedded in GaAsSb layers, it has been reported that the type-I/type-II transition occurs at an Sb composition of 14%.4,5In addition to the application in QDs laser, the type-II InAs/ GaAsSb QDs heterostructures are potentially useful materials for photovoltaic devices due to their nature of the separation of photoexcited electrons and holes.4–7 InAs/ GaAsSb QDs system are therefore very attractive for both of academic interest and industrial application. How-ever, the published reports on optical properties of this ma-terial system are still rather limited. Here, we present a de-tailed investigation of type-II InAs/ GaAs0.7Sb0.3 QDs by photoluminescence共PL兲 measurement. It is found that a sig-nificant blueshift of the peak energy occurs under a moderate excitation power, which can be well explained by the band bending effect due to the spatially separated photoexcited carriers in type-II band alignment. In addition, a strong op-tical anisotropy is also found, which is attributed to the asymmetry of interface chemical bonds. In view of the ma-ture growth process, InAs/ GaAsSb QDs should be able to serve as a model system for the search of novel properties in type-II quantum structures.

The sample was grown in Stranski-Krastanov 共SK兲 growth mode on an undoped semi-insulating GaAs 共100兲 substrate by a VG V-80MKII solid-source molecular beam epitaxy machine.8 After a 500 nm thick undoped GaAs buffer layer, the sample consists of a 2.0 ML InAs QDs

embedded between GaAs0.7Sb0.3 layers grown at 485 ° C, then a 60 nm undoped GaAs without growth interrupt, and a 500 nm undoped GaAs was grown at 500 ° C. The antimony composition in the GaAs0.7Sb0.3layer was determined using double crystal x-ray diffraction 共DXRD兲 measurement. In addition, a sample containing a single layer of InAs/ GaAs QDs共2.0 ML兲 was used for comparison.

For the structural measurement, transmission electron mi-croscopy共TEM兲 and scanning transmission electron micros-copy共STEM兲 studies were performed using a Philips Tecnai-F20 instrument. The scanning electron microscopy 共SEM兲 images were recorded by a JEOL-JSM 6500 system. The PL spectra were measured by a Spectra Pro 300i monochro-mator and an InGaAs detector. An Ar-ion laser working at 514.5 nm was used as the excitation source. For the mea-surement of the polarization dependence of the PL spectra, a depolarizing filter was placed in front of the spectrometer to exclude the possible polarized character of the grating.

Figures 1共a兲 and 1共b兲 show the SEM images of the un-capped InAs QDs deposited on GaAs and GaAs0.7Sb0.3 lay-ers, respectively. The size of the uncapped InAs QDs formed on GaAs is smaller than those on GaAs0.7Sb0.3, which is caused by the different surface strain field as reported previously.9,10 Figure 1共c兲 clearly shows the 关100兴-axial cross-sectional TEM image of the InAs QDs embedded be-tween GaAs0.7Sb0.3 layers with a total thickness of about 11 nm as marked by the separation of the two dash lines. But the actual shape of the InAs QD is unclear since the strain field may extend beyond the island boundary. We thus have performed the STEM measurement, which is known not to be influenced by the strain field. From the STEM image, the InAs QD is determined to be about 3.5 nm high and reveals a flat top as shown in Fig. 1共d兲. The combination of TEM and STEM images therefore clearly shows that the InAs QD

PHYSICAL REVIEW B 75, 033310共2007兲

(2)

is indeed sandwiched by GaAsSb layers. Each GaAsSb layer has a thickness of about 5.5 nm.

Figure 2 shows the PL spectra of InAs/ GaAs and InAs/ GaAs0.7Sb0.3 QDs under different excitation power at 20 K. The peak energy of InAs/ GaAs0.7Sb0.3QDs is smaller than that of InAs/ GaAs QDs at the same excitation power. Quite interestingly, in spite of having a type-II band align-ment, the InAs/ GaAs0.7Sb0.3QDs exhibit a strong emission with intensity close to that of the InAs/ GaAs QDs at the same excitation condition. Besides, the linewidth of the emission spectra of the InAs/ GaAs0.7Sb0.3 QDs is about 50 meV which is smaller than that of the analogous system of InGaAs QDs embedded between GaAsSb layers as re-ported previously.6 Both of the bright emission and narrow linewidth indicate that the sample has an excellent quality. As shown in Fig.3, in contrast to the unchanged peak energy of type-I InAs/ GaAs QDs, we observe a giant blueshift of about 29 meV in InAs/ GaAs0.7Sb0.3 QDs under a moderate excitation power. The value of the blueshift of the

InAs/ GaAsSb QDs is close to 34 meV of the analogous structure in a previous report.5This peculiar behavior cannot be due to laser heating because it will cause a redshift in the band gap. The blueshift is too large to be caused by the state filling of the localized states due to interface roughness or alloy potential fluctuations.11We believe that the underlying mechanism of our observation is an intrinsic nature of the type-II band alignment of InAs/ GaAs0.7Sb0.3. Figure 4 shows the schematic type-II band diagram for InAs/ GaAs0.7Sb0.3system with the band bending effect due to charge carriers transfer. The electrons and holes are con-fined in the InAs QDs and GaAs0.7Sb0.3layer, respectively.4,5 The giant blueshift under a moderate optical excitation level can now be readily understood.12 Qualitatively, it can be explained as follows. The spatially separated electrons and holes lead to the appearance of a strong electric field at the interface which in turn gives rise to the bending of the valence and conduction band. With increasing excitation power, the band bending effect becomes more pronounced,

FIG. 1. Scanning electronic microscopy images of uncapped共a兲 InAs/ GaAs and共b兲 InAs/GaAs0.7Sb0.3quantum dots.共c兲 Transmis-sion electron microscopy and 共d兲 scanning transmission electron microscopy images of the InAs quantum dots embedded between GaAs0.7Sb0.3layers.

FIG. 2. Excitation power dependence of the photoluminescence spectra.

FIG. 3. Peak energy of photoluminescence spectra for InAs/ GaAs共䊏兲 and InAs/GaAs0.7Sb0.3共䉱兲 recorded with different

excitation level. The inset shows the peak energy of

InAs/ GaAs0.7Sb0.3共䉱兲 on the cubic root of excitation power 共solid line兲.

FIG. 4. Band-bending effect of the type-II band structure under low共solid line兲 and high 共dash line兲 excitation power density.

BRIEF REPORTS PHYSICAL REVIEW B 75, 033310共2007兲

(3)

and the accumulated electrons and holes will be confined in a narrower region near the interface. Consequently, the elec-tron quantization energy is enhanced and the blueshift oc-curs. As the excitation power exceeds about 100 mW, the blueshift saturates due to the filling effect of energy states.

In order to analyze the blueshift quantitatively as shown in Fig.3, we consider the change of the nonequilibrium rier density due to photoexcitation. The strong localized car-riers near the interface form a charged plane, and corre-spondingly, produce an approximated triangular potential well with an electric field of␧⬀I1/2, where I is the excitation photon flux. The ground electron state Ecin a triangular well is given by8,12

Ec= const ·␧2/3= const

· I1/3. 共1兲 The electron quantization energy is thus expected to increase proportionally with the third root of the excitation power. As shown in the inset of Fig. 3, the PL peak energy can be described quite well by the third root of the excitation power obtained from Eq.共1兲. It indicates that the giant blueshift is indeed a peculiar property of the type-II heterostructure.

Besides, with increasing the excitation power, the line shape of type-I InAs/ GaAs QDs remains unchanged as shown in Fig.2. In contrast, the evidently asymmetric spec-tra of InAs/ GaAs0.7Sb0.3 at low excitation power gradually turns into a symmetric shape. This behavior may be attrib-uted to the fact that the built-in electric field due to surface charges can tilt the band edge, which will give rise to a nonuniform distribution of the photoexcited carriers in con-duction and valence bands. Therefore, the line shape be-comes asymmetric and results in the so-called quantum con-finement Stark effect共QCSE兲.13,14With increasing excitation power, the internal electric field is screened by the photoex-cited carriers, and therefore the linewidth gradually turns into a symmetric shape.

Additionally, we have performed the polarization depen-dence of PL spectra for InAs/ GaAs and InAs/ GaAs0.7Sb0.3 QDs, as shown in Fig.5. In bulk semiconductors, the study of polarization dependent PL spectra is an efficient tool for obtaining the information about symmetry of structures and

emission states. In semiconductor QDs, the bulk valance band is split into heavy-hole, light-hole, and spin-orbit band via quantum confinement and uniaxial strain.15,16These three bands have different polarization properties, in general, the lowest optical transition are related to geometric symmetries of the QDs. Indeed, for the type-I InAs/ GaAs QDs, the iso-tropic emission light is consistent with the isoiso-tropic shape of uncapped QDs, as shown in Fig.1共a兲. Quite surprisingly, the InAs/ GaAs0.7Sb0.3QDs exhibit a strong optical anisotropy as shown in Fig. 5. The polarization degree defined as P ⬅共Imax− Imin兲/共Imax+ Imin兲 is as large as 24%, where the Imax and Imin are the maximum and minimum polarized emission intensities, respectively. From the SEM image as shown in Fig. 1共b兲, the uncapped InAs QDs grown on GaAs0.7Sb0.3 layer do not reveal an evidently anisotropic shape. Therefore, the observed optical anisotropy cannot be due to the geomet-ric effect of the QDs. Besides, this behavior also cannot be attributed to a slight misorientation of the substrate which may give rise to a preferential orientation of steps at the interface. It is because both samples in our studies are grown on the same kind of substrate, and the InAs/ GaAs QDs do not show any indication of anisotropic behavior.

We believe that the behavior arises from the type-II tran-sition for electrons confined in the InAs QDs and holes con-fined in the GaAs0.7Sb0.3layer, in which the recombination is due to the electron-hole wave function overlap across the interface within an extremely narrow region. Therefore, the transition oscillator strength is strongly correlated with the anisotropic property of interface chemical bonds.17–19 Com-pared with the interior volume of a bulk material, a lower symmetry does exist in an isolated interface between two semiconductors. The resultant optical anisotropy has been observed previously in many semiconductor heterostructures with an AB/ABC-type combination.18,19 More specifically, the in-plane anisotropy arises from the uncompensated polar-ization of chemical bonds across the interface, not found inside the bulk. A schematic diagram of the polarized chemi-cal bonds for a GaAsSb layer capped on a InAs QD is shown in Fig. 6. In the zinc-blende structure, the interface of the

FIG. 5. Polar plot of the polarized photoluminescence intensity as a function of the polarizer angle for InAs/ GaAs 共䊏兲 and InAs/ GaAs0.7Sb0.3共䉱兲 quantum dots. The dash curve is a cos2␪ fit.

FIG. 6. Schematic diagram of the bond sequence of

InAs/ GaAs0.7Sb0.3quantum dots.

BRIEF REPORTS PHYSICAL REVIEW B 75, 033310共2007兲

(4)

InAs/ GaAsSb heterostructure consists of In-Sb, In-As, and As-Ga bonds. We can clearly see that the chemical bonds are strongly anisotropic, which is quite different from the bulk case. Therefore, the in-plane anisotropic characteristic inher-ently exists in the InAs/ GaAs0.7Sb0.3 QDs studied here. To verify the anisotropic effect, we have rotated InAs/ GaAs0.7Sb0.3 QDs sample by 90°, and found that the direction of the polarized emission also rotates by 90°. This result provides a further evidence to support the fact that the optical anisotropy does depend on the orientation of interface chemical bonds.

In order to clarify that the observed polarized PL spectra are induced by the anisotropic nature of the interface chemi-cal bonds, we have performed the excitation power and tem-perature dependences of the polarization. It is found that the degree of polarization is insensitive to the change of excita-tion power in the range from 10 to 200 mW. The degree of polarization is also very stable with respect to the change of temperature from 10 to 200 K. These results can be used to rule out extrinsic mechanisms related to the in-plane aniso-tropy. For example, the built-in electric fields caused by un-intentional doping will be screened under light irradiation. We can also exclude a significant role of localized states and nonradiative channels in the formation of the in-plane aniso-tropy. Since they will be gradually saturated by the excitation source and the thermally activated carriers will redistribute among them, which will change the transitions of carriers. We thus conclude that the polarization of the spatially

indi-rect PL in InAs/ GaAs0.7Sb0.3 is an inherent nature of the interface chemical bonds. Polarized PL measurements there-fore provide a simple tool to probe interface anisotropy in semiconductor heterostructures.

In summary, the optical properties of type-II InAs/ GaAs0.7Sb0.3QDs have been investigated by PL mea-surements. In contrast to the InAs/ GaAs QDs, it is found that the PL spectra show a giant blueshift under a moderate optical excitation level. This behavior can be interpreted in terms of the band bending effect due to the spatially photo-excited carriers in a type-II band alignment. We also found that the PL spectra exhibit a large in-plane polarization with the polarization degree up to 24%. The large polarization does not depend on the excitation power as well as tempera-ture, which can be used to exclude the possibility of extrinsic mechanisms related to the in-plane anisotropy. The observed optical anisotropy of InAs/ GaAs0.7Sb0.3 QDs can be ex-plained quite well in terms of the wave function overlap between electrons and holes across the heterostructure inter-face. Therefore, the emission reflects the intrinsic property of the orientation of chemical bonds at the type-II heterointer-faces. Our results shown here should be very useful to serve as a starting point for the study of the optical properties of other type-II semiconductor quantum dots.

This work was supported by the Education of Ministry and National Science Council of the Republic of China.

*Corresponding author. Email address: [email protected]共Y. F. Chen兲.

1H. Y. Liu, C. M. Tey, I. R. Sellers, T. J. Badcock, D. J. Mowbray, M. S. Skolnick, R. Beanland, M. Hopkinson, and A. G. Cullis, J. Appl. Phys. 98, 083516共2005兲.

2S. Anantathanasarn, R. Nötzel, P. J. van Veldhoven, F. W. M. van Otten, T. J. Eijkemans, and J. H. Wolter, Appl. Phys. Lett. 88, 063105共2006兲.

3Hideaki Saito, Kenichi Nishi, and Shigeo Sugou, Appl. Phys. Lett. 73, 2742共1998兲.

4H. Y. Liu, M. J. Steer, T. J. Badcock, D. J. Mowbray, M. S. Skolnick, F. Suarezb, J. S. Ng, M. Hopkinson, and J. P. R. David, J. Appl. Phys. 99, 046104共2006兲.

5H. Y. Liu, M. J. Steer, T. J. Badcock, D. J. Mowbray, M. S. Skolnick, P. Navaretti, K. M. Groom, M. Hopkinson, and R. A. Hogg, Appl. Phys. Lett. 86, 143108共2005兲.

6J. M. Ripalda, D. Granados, Y. González, A. M. Sánchez, S. I. Molina, and J. M. García, Appl. Phys. Lett. 87, 202108共2005兲. 7Kouichi Akahane, Naokatsu Yamamoto, Shin-ichiro Gozu, and

Naoki Ohtani, Physica E共Amsterdam兲 26, 395 共2005兲. 8D. Bimberg, M. Grundmann, and N. N. Ledentsov, Quantum Dot

Heterostructures共Wiley, New York, 1999兲.

9J. P. McCaffrey, M. D. Robertson, P. J. Poole, B. J. Riel, and S. Fafard, J. Appl. Phys. 90, 1784共2001兲.

10Weihong Jiang, Huaizhe Xu, Bo Xu, Wei Zhou, Qian Gong, Ding

Ding, Jiben Liang, and Zhanguo Wang, J. Vac. Sci. Technol. B 19, 197共2001兲.

11J. M. Iannelli, J. Maserjian, B. R. Hancock, P. O. Andersson, and F. J. Grunthaner, Appl. Phys. Lett. 54, 301共1989兲.

12Y. S. Chiu, M. H. Ya, W. S. Su, and Y. F. Chen, J. Appl. Phys. 92, 5810共2002兲.

13A. I. Yakimov, A. V. Dvurechenskii, A. I. Nikiforov, V. V. Uly-anov, A. G. Milekhin, A. O. Govorov, S. Schulze, and D. R. T. Zahn, Phys. Rev. B 67, 125318共2003兲.

14T. T. Chen, C. H. Chen, W. Z. Cheng, W. S. Su, M. H. Ya, Y. F. Chen, P. W. Liu, and H. H. Lin, J. Appl. Phys. 93, 9655共2003兲. 15M. Henini, S. Sanguinetti, S. C. Fortina, E. Grilli, M. Guzzi, G. Panzarini, L. C. Andreani, M. D. Upward, P. Moriarty, P. H. Beton, and L. Eaves, Phys. Rev. B 57, R6815共1998兲.

16P. Yu, W. Langbein, K. Leosson, J. M. Hvam, N. N. Ledentsov, D. Bimberg, V. M. Ustinov, A. Yu. Egorov, A. E. Zhukov, A. F. Tsatsul’nikov, and Yu. G. Musikhin, Phys. Rev. B 60, 16680 共1999兲.

17A. V. Platonov, V. P. Kochereshko, E. L. Ivchenko, G. V. Mikhailov, D. R. Yakovlev, M. Keim, W. Ossau, A. Waag, and G. Landwehr, Phys. Rev. Lett. 83, 3546共1999兲.

18J. Böhrer, A. Krost, R. Heitz, F. Heinrichsdorff, L. Eckey, D. Bimberg, and H. Cerva, Appl. Phys. Lett. 68, 1072共1996兲. 19Y. S. Chiu, M. H. Ya, W. S. Su, T. T. Chen, Y. F. Chen, and H. H.

Lin, Appl. Phys. Lett. 81, 4943共2002兲.

BRIEF REPORTS PHYSICAL REVIEW B 75, 033310共2007兲

數據

FIG. 3. Peak energy of photoluminescence spectra for InAs/ GaAs 共䊏兲 and InAs/GaAs 0.7 Sb 0.3 共䉱兲 recorded with different excitation level

參考文獻

相關文件

Rather than requiring a physical press of the reset button before an upload, the Arduino Uno is designed in a way that allows it to be reset by software running on a

As the result, I found that the trail I want can be got by using a plane for cutting the quadrangular pyramid, like the way to have a conic section from a cone.. I also found

In part II (“Invariance of quan- tum rings under ordinary flops II”, Algebraic Geometry, 2016), we develop a quantum Leray–Hirsch theorem and use it to show that the big

It is well known that second-order cone programming can be regarded as a special case of positive semidefinite programming by using the arrow matrix.. This paper further studies

• We have found a plausible model in AdS/CFT that captures some essential features of the quantum hall eect as well as a possible experimental prediction. • Even if this is the

• A way of ensuring that charge confinement does occurs is if there is a global symmetry which under which quarks (heavy charges) are charged and the gluons and other light fields

S15 Expectation value of the total spin-squared operator h ˆ S 2 i for the ground state of cationic n-PP as a function of the chain length, calculated using KS-DFT with various

The temperature angular power spectrum of the primary CMB from Planck, showing a precise measurement of seven acoustic peaks, that are well fit by a simple six-parameter