phys. stat. sol. (c) 1, No. 3, 581 – 584 (2004) / DOI 10.1002/pssc.200304044

Hydrogenation of strain engineered InAs/InxGa1–xAs quantum dots D. Ochoa1, A. Polimeni1, M. Capizzi*, 1, A. Frova1, L. Seravalli2, M. Minelli2 , P. Frigeri2, and S. Franchi2 1 2

INFM – Physics Dept., Univ. of Roma “La Sapienza”, P.le A. Moro 2, I-00185 Roma, Italy CNR-IMEM Institute, Parco delle Scienze, 37a, I-43100 Parma, Italy

Received 15 September 2003, accepted 18 September 2003 Published online 2 February 2004 PACS 71.20.Nr, 78.55.Cr, 78.66.Fd, 81.05.Ea, 81.15.Hi InxGa1–xAs/InAs/InxGa1–xAs structures have been grown by atomic layer molecular beam epitaxy on top of a GaAs buffer and substrate. In these structures, the thickness d and/or the In composition x of the lower InxGa1–xAs confining layer control the strain in self-assembled quantum dots. This strain engineering has allowed achieving emission energies as low as 1.5 µm at low T, with a rapid quenching of the photoluminescence (PL) signal at high T. Hydrogen irradiation of these structures leads to an increase in the PL efficiency, higher in samples with higher x, with a blue shift in the peak energy. A higher concentration of non radiative defects in confining layers richer in indium is responsible for the observed PL quenching, more than an increased thermal escape of carriers toward the InxGa1–xAs barriers. © 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Introduction Fundamental properties and technological applications of self-assembled quantum-dots (QDs) account for the great interest for these nanostructures [1]. Since QD may self-assemble on a GaAs substrate, efficient vertical cavity surface-emitting QD lasers with low and temperature-independent threshold current densities can be obtained by using AlxGa1–xAs/GaAs-based Bragg reflectors [2, 3]. A great effort has been devoted, therefore, to achieve room temperature (RT) QD light emission at 1.3 and 1.55 µm, the transmission windows of silica optical fibers. In the case of InAs QDs, this goal has been pursued increasing the QD size by using atomic layer molecular-beam epitaxy [4] (ALMBE) growth, low deposition rates [5], activated alloy phase separation [6], and by reducing the strain with upper confining layers (UCLs) [7] or with both UCLs and lower confining layers (LCLs) of InxGa1–xAs [8]. Recently, the role of strain and carrier confinement potentials on the QD photoluminescence (PL) emission energy has been investigated in QDs deposited on LCLs with different thicknesses, d, and given In compositions, x. In this way, the strain of QDs could be controlled independently of In composition in confining layers [9]. This QD strain engineering has allowed achieving RT emission at 1.3 µm from InAs/InxGa1-xAs QDs with x = 0.15 and d = 60 nm [9]. Moreover, low temperature emission energy as low as 1.50 µm has been achieved in structures with x = 0.35 and d = 500 nm. However, emission at and beyond 1.55 µm could not be reached at RT because of a rapid quenching of the PL intensity with increasing T [9]. This quenching may be due to an increased carrier thermal escape toward the InxGa1-xAs barrier, which is related to a reduced band-gap discontinuity, and/or to an exceedingly high concentration of non radiative defects in LCLs with high indium concentration. The role of non radiative defects in the PL thermal quenching has been investigated by irradiating with hydrogen QD structures like those studied in Ref. 9. Hydrogen irradiation leads usually to an efficient passivation of point defects in semiconductors [10], giving rise, in turns, to orders of magnitude *

Corresponding author: e-mail: [email protected], Phone: +39 06 4991 4381, Fax: +39 06 4991 4387 © 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

582

D. Ochoa et al.: Hydrogenation of strain engineered InAs/InxGa1–xAs quantum dots

increase in the PL efficiency of low-T grown, highly defected QD precursors [11]. At a H dose dH = 1·1018 ions/cm2, all samples investigated here show an increase in the PL efficiency. This increase is maximum for samples with the highest indium concentration in the LCLs, where a higher concentration of structural defects is expected due to the larger lattice mismatch to the GaAs buffer and substrate. This PL increase allows also increasing sizably the highest QD emission temperature. Therefore, mainly non radiative defects account for the observed thermal quenching. At a H dose five times higher, the insurgence of new H induced defects leads to a decrease in the PL efficiency in the samples with lowest x. For high x samples, instead, the PL peak emission keep increasing and shifts surprisingly to higher energy. Experimental details and data The structures investigated here consist of 1) a 100-nm-thick GaAs buffer layer grown by molecular beam epitaxy, MBE, on a (100) GaAs substrate; 2) a InxGa1-xAs LCL with thicknesses, d, ranging from 60 nm to 500 nm, grown by MBE at 490 °C; 3) a plane of InAs QDs with a 3-monolayer (ML) coverage deposited by ALMBE [4] at 460 °C; 4) a 20-nm-thick InxGa1-xAs UCL grown by ALMBE at low temperature (360 °C) to reduce the interaction among confining layers and QDs. Before and after the QD deposition, the growth was interrupted for 210 s to change substrate temperature. The In composition in confining layers was 0.15, 0.28, 0.31, and 0.35. Hydrogen irradiation was performed post growth using a Kaufmann source with the samples maintained at 300 °C. Samples were excited with the 532 line of a vanadate-YAG laser. PL spectra were taken at 10 K, dispersed by a single 0.75 m monochromator and detected by a LN-cooled (InGa)As linear array (0.2 meV spectral resolution). Photoluminescence spectra taken at 10 K in as grown (continuous lines), lightly hydrogenated (dashed lines, dH = 1x1018 ions/cm2), and highly hydrogenated (dotted lines, dH = 5x1018 ions/cm2)

Fig. 1 QD photoluminescence spectra at 10 K for different H doses, dH, in InAs/InxGa1-xAs heterostructures with different In concentration, x, and thickness of the lower confining layer, d. Continuous, dashed, and dotted lines refer to spectra in as grown, hydrogenated at dH = 1x1018 cm-2, and hydrogenated at dH = 5x1018 cm-2 samples, respectively.

representative samples are shown in Fig. 1. Multiplicative factors refer always to as grown samples, whose PL spectra (continuous lines) we discuss first. The x = 0.15, d = 220 nm sample shows a PL band with a single peak at 1250 nm, which shifts to 1350 nm at RT with a still sizable emission; see Ref. 9. Also the x= 0.31, d = 220 nm sample shows a single peaked (1360 nm) PL band, which, however, quenches already at 190 K. The x = 0.35, d = 220 nm sample, whose PL emission is peaked at ~1500 nm, shows a quite structured PL band whose intensity vanishes at about 120 K. Measurements as a function of power density and temperature, not shown here, indicate that the structures of the x = 0.35 sample are due to different QD families. Topographic images of structures similar to those investigated here but grown without UCLs were taken by contact-mode atomic force microscopy (AFM). These AFM measurements show that the average QD size increases with x, thus accounting at least in part for the con© 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

phys. stat. sol. (c) 1, No. 3 (2004) / www.pss-c.com

583

comitant decrease of the QD emission band peak energy [9], and suggest that QD size distribution widens also with x. The decrease with In content in the sample emission efficiency indicates that the concentration of non radiative defects increases with x, which leads to a decreasing quenching temperature. In the samples irradiated at dH = 1x1018 ions/cm2 (dashed lines), the radiative emission integrated over the whole QD band increases with respect to that of as grown samples by a factor of about 5, except for an x = 0.35, d = 500 nm sample where this factor is equal to 15 (not shown here). An increase in the PL intensity of a factor of 5 has been also observed at RT in the x = 0.15 sample.

Fig. 2 Dependence on temperature of the PL intensity in two representative samples before and after H irradiation.

The effects of increasing the H dose to dH = 5x1018 ions/cm2 are shown in Fig. 1 by the dotted curves. The PL intensity decreases drastically with respect to the as grown sample in the x = 0.15, d = 220 nm sample, while it keeps increasing in the other samples with higher In concentration. The PL peak energy, instead, exhibits a blue shift that increases with x and d (going from ~30 meV in the x = 0.15, d = 220 nm sample to ~60 meV in the x = 0.31and 0.35 nm samples). The dependence of PL integrated intensity on H dose can be accounted for by an increasing density of non radiative defects with x, already invoked to explain the decrease with x of the emission efficiency in the as grown samples. Moreover, in the case of the x = 0.15, d = 220 nm sample, which has the highest luminescence efficiency among the as grown samples, one should invoke the insurgence of new H induced defects at high H dose when all non radiative native defects have been passivated, as already observed previously [11]. The change in the PL lineshape at the highest dH, instead, is quite unexpected and surprising. On a purely phenomenological ground, it could be attributed to the insurgence of H induced non radiative defects more pronounced in the case of larger QDs. A more extensive study of the H induced PL changes should be accompanied by AFM measurements in order to investigate this new phenomenon in some detail. Finally, as a consequence of the increase in the PL efficiency with hydrogenation, the temperature where a sizable QD emission is still observed increases from 190 K to RT in samples with x = 0.31, and from 120 K to 220 K in the sample with x = 0.35, as shown in Fig. 2 for dH = 5x1018 ions/cm2. It should be noticed that the PL band lineshape does not change much with temperature in the as grown samples, while its center of mass moves to lower energy for increasing temperature in the hydrogenated samples. This provides an additional evidence that non radiative centers dominate the T dependence in the former samples, while carrier localization effects are dominant in the latter samples where H irradiation leads to the passivation of most non radiative centers. In conclusion, strain engineered InAs/InxGa1-xAs heterostructures have been irradiated with hydrogen. PL measurements have shown that the quenching of the QD emission with temperature reported in these structures is mainly due to non radiative defects. These defects can be passivated by hydrogen thus increasing the maximum temperatures at which a sizable PL signal is measured. More extensive and de© 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

584

D. Ochoa et al.: Hydrogenation of strain engineered InAs/InxGa1–xAs quantum dots

tailed investigation of this effect are in progress in order to optimize the post growth H treatments and understand the nature of a blue shift observed in the PL peak maximum at high H doses. Acknowledgements This work has been funded by the European Commission GROWTH program, within the framework of the NANOMAT project, Contract no. G5RD-CT-2001-00545..

References [1] D. Bimberg, M. Grundmann, and N. N. Ledentsov, Quantum Dot Heterostructures (Wiley, Chichester, 1999). [2] M. Asada, Y. Miyamoto, and Y. Suematsu, IEEE J. Quantum Electron. 22, 1915 (1986). [3] N. N. Ledentsov, D. Bimberg, V. M. Ustinov, Zh. I. Alferov, and J. A. Lott, Physica E (Amsterdam) 13, 871 (2002). [4] R. P. Mirin, J. P. Ibbetson, J. E. Bowers, and A. C. Gossard, J. Cryst. Growth 175/176, 696 (1997); A. Bosacchi, P. Frigeri, S. Franchi, P. Allegri, and V. Avanzini, J. Cryst. Growth 175/176, 771 (1997). [5] P. B. Joyce, T. J. Krzyzewski, G. R. Bell, T. S. Jones, R. Murray, D. Childs, and S. Malik, Phys. Rev. B 62, 10891 (2000). [6] M. V. Maximov, A. F. Tsatsul’nikov, B. V. Volovik, D. S. Sizov, Yu. M. Shernyakov, I. N. Kaiander, A. E. Zhukov, A. R. Kovsh, S. S. Mikhrin, V. M. Ustinov, Zh. I. Alferov, R. Heitz, V. A. Shchukin, N. N. Ledentsov, D. Bimberg, Yu. G. Musikhin, and W. Nuemann, Phys. Rev. B 62, 16671 (2000). [7] H. Nishi, H. Saito, S. Sugou, and J. S. Lee, Appl. Phys. Lett. 74, 1111 (1999). [8] A. Stintz, G. T. Liu, A. L. Gray, R. Spillers, S. M. Delgado, and K. J. Malloy, J. Vac. Sci. Technol. B 18, 1496 (2000). [9] L. Seravalli, M. Minelli, P. Frigeri, P. Allegrini, V. Avanzini, and S. Franchi, Appl. Phys. Lett. 82, 2341 (2003). [10] Jacques I. Pankove and Nobel M. Johnson (eds.), Hydrogen in Semiconductors, edited by R.K. Willardson and Albert C. Beer, Semiconductors and Semimetals, Vol. 34 (Academic, New York, 1991). [11] A. Polimeni, D. Marangio, M. Capizzi, A. Frova, and F. Martelli, Appl. Phys. Lett. 65, 1254 (1994).

© 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim