GaAs core / SrTiO 3 shell nanowires grown by Molecular Beam Epitaxy

Transcription

1 Supporting Information for GaAs core / SrTiO 3 shell nanowires grown by Molecular Beam Epitaxy X. Guan 1, J. Becdelievre 1, B. Meunier 1, A. Benali 1, G. Saint-Girons 1, R. Bachelet 1, P. Regreny 1, C. Botella 1, G. Grenet 1, N. P. Blanchard 2, X. Jaurand 3, M. G. Silly 4, F. Sirotti 4, N. Chauvin 5, M. Gendry 1, J. Penuelas 1 1 Institut des Nanotechnologies de Lyon - Université de Lyon, UMR CNRS, Ecole Centrale de Lyon, 36 avenue Guy de Collongue, F Ecully cedex, France 2 Institut Lumière Matière (ILM), UMR5306 Université Lyon 1- CNRS Université de Lyon, Villeurbanne cedex, France 3 Centre Technologique des Microstructures, Université Claude Bernard Lyon1, 5 rue Raphael Dubois-Bâtiment Darwin B, F-69622, Villeurbanne Cedex, France 4 Synchrotron SOLEIL (TEMPO beamline), l Orme des Merisiers, Saint-Aubin, Gifsur-Yvette, France 5 Institut des Nanotechnologies de Lyon - Université de Lyon, UMR CNRS, INSA- Lyon, 7 avenue Jean Capelle, Villeurbanne, France 1

2 1. Sample preparation and effect of the growth temperature: 2. Shadowing: 3. GaAs / SrTiO 3 interface and SrTiO 3 structural quality: 4. X-ray photoelectron spectroscopy analysis and thermal stability: 5. Photoluminescence measurements 2

3 1. Sample preparation and effect of the growth temperature: GaAs NWs were grown on a cm 2 Si (111) substrate. After As capping, this sample was cut into small pieces which were inserted into another sample holder dedicated to the oxide MBE reactor in order to avoid cross contaminations. Figure S1: EDS pattern measured on a single GaAs / SrTiO 3 NW. Figure S1 shows the EDS pattern of a single GaAs / SrTiO 3 NW measured after scrapping the NWs on a TEM grid. No contaminations are visible. The Cu peaks arise from the TEM grid. In order to investigate the impact of the growth parameter on the structural and morphological properties a series of samples was prepared. The one-step deposition of 40 nominal monolayers of SrTiO 3 under low O 2 partial pressure of 5x10-8 Torr was performed for comparing with the as-mentioned two-step method. Using this one-step method, we prepared three different GaAs / SrTiO 3 core / shell NW arrays by only modifying the SrTiO 3 deposition temperature. 3

4 Figure S2: TEM images of GaAs / SrTiO 3 grown at 350 C (a-c). STEM image of a typical NW (d). EDS line scan corresponding to the marked white line (e). Figure S2 shows TEM measurements performed on the sample for which the SrTiO 3 was grown at 350 C. At the beginning of the growth, the RHEED pattern was consistent with GaAs nanowires, during the SrTiO 3 growth, the spot intensity decreased slowly in accordance with an amorphous layer covering the GaAs nanowires as expected at such low temperature. As shown by the low magnification TEM image (Figure S2.a), the newly fabricated surface of the core / shell NW is rather smooth and the SrTiO 3 layer is quite uniform. However in HRTEM images, besides amorphous layer (Figure S2.b), some crystallites of SrTiO 3 are also observed (Figure S2.c). An example is shown in figure S2.c where the observed lattice is ascribed to (110) SrTiO 3 based on the FFT analysis (Figure S2.c insert). However, there is no particular epitaxial relationship between the SrTiO 3 shell and the GaAs core. Either the SrTiO 3 has started to be crystalized at such low temperature, or some parts of the shell are very sensitive to the electron beam during the TEM measurement. Finally, for confirming the 4

5 Supporting Information core shell hetero-structure, EDS line scan (Figure S2.e) has been performed, corresponding to the marked white line in the STEM image of one typical core / shell NW (Figure S2.d). Ga and As exist only in the core, while Sr and Ti are distributed in the shell. Figure S3: TEM images of GaAs / SrTiO3 grown at 450 C (a-c). Electron diffraction pattern (d) corresponding to the region shown in (c). The second NW array was obtained by depositing SrTiO3 layer at 450 C under the same oxygen partial pressure. As shown in TEM images (Figure S3.a) a lot of small particles appear in the NW surface, making it uneven, while some parts of the SrTiO3 layer still maintain the amorphous phase (Figure S3.b). However, the shell hasn t shown hints of epitaxial growth. Rings in accordance with polycrystalline SrTiO3 are evidenced in the electron diffraction pattern (Figure S3.d) taken from the region displayed in Figure S3.c. 5

6 Figure S4: TEM images for different magnifications of the side surface of a typical GaAs / SrTiO 3 NW grown at 620 C (a-c). Then, the SrTiO 3 growth temperature was increased to 620 C for the last sample. As expected, the SrTiO 3 layer is well crystalized. Small pyramidal-shaped SrTiO 3 crystals are formed at the surface of the GaAs NW core, giving the core / shell NW a zigzag surface (Figure S4). FFT process has been performed for the marked zone (Figure S4.c), typical cubic pattern has been obtained, interplannar spacing 2.76 Å is attributed to SrTiO 3 {110}. The relative orientation of this crystal and the GaAs NW core obtained directly from Figure S4.c is [111]GaAs // [110]SrTiO 3. Nevertheless, it is only one special case, the majority of SrTiO 3 6

7 crystals were grown randomly without one favourable orientation, such as the particle labelled by the blue arrow in Figure S4.c. In summary, deposition at high temperature promotes the SrTiO 3 crystallization with some epitaxial crystallites while induces the formation of dewetted islands and avoids the formation of a continuous shell that perfectly wraps the NW core. The two-step SrTiO 3 growth method allows obtaining a relatively good wetting due to the low deposition temperature of the first step while the second step (an annealing followed by high temperature SrTiO 3 growth) allows obtaining the SrTiO 3 perovskite crystalline structure. 2. Shadowing: Figure S5: EDS line scan performed along six different GaAs / SrTiO 3 NWs. Figure S5 shows EDS line scans measured along the direction perpendicular to the NW growth direction for six different NWs. Sr, Ti and O are systematically detected. In most of the cases the line scans are in accordance with core shell morphology, i.e. Sr, Ti and O wrap 7

8 the GaAs NW core. However, in some cases an asymmetry of the measured amount of Sr, Ti and / or O is detected (Figure S5.a,d). In figure S5.d an entire GaAs side is not covered by SrTiO 3 which could be the consequence of shadowing effects: The self-shadowing effect has been largely reported in the literature. Shadowing has been reported when: - The incident angle of the deposition flux is not perpendicular to the substrate surface [1]. - NWs are not verticals, so some facets are not directly exposed to the molecular beam. - The incident beam flux is intercepted only by the upper part of NWs (base shadowing) [2,3] - Lengths of NWs are not uniform, so shorter NWs are shadowed by longer neighbours [4,5] - The arrival of vapour atoms at the sidewalls may be restricted due to the high wire densities [5,6] In order to avoid the self-shadowing effect, several parameters must be adjusted as the deposition rate [7], the incidence angle of the molecular beam flux [7], the substrate rotation speed [7, 8], the density of NWs by lithography and precisely spreading catalytic droplets [9, 10] and the NW verticality. In our case, the most probable reasons for the shadowing effect could be that few NWs are not verticals and despite the low NW density some NWs appear to be very close from each others. 3. GaAs / SrTiO 3 interface and SrTiO 3 structural quality: 8

9 Figure S6: Cross section of the RHEED pattern along the normal to the surface at the end of the SrTiO 3 growth. As shown in previous studies the epitaxial growth of SrTiO 3 on Si (001) or GaAs (001) substrates is mainly driven by the minimization of the lattice mismatch leading to the alignment of SrTiO 3 [100] with Si [1-10] axis [11-13] or with GaAs axis [1-10] [14,15]. Unfortunately, others substrate orientations have not been intensively studied. Gao et al. reported on the growth of (110) oriented SrTiO 3 crystal on Si(001) [16] and also on Si(111) and Si(110) [17]. To our knowledge there is no study reporting on the SrTiO 3 growth on GaAs (110) or (111) substrates. Figure S7 shows a possible model of the interface between SrTiO 3 nanocrystals and GaAs (1-10) facet (in accordance with Figure 4.b). The interface of Figure S7 corresponds to the following in plane alignment ( in plane referring to the crystallographic axis parallel to the facets and out of plane referring to the direction perpendicular to the facets): SrTiO 3 [001] // GaAs [110] and SrTiO 3 [1-10] // GaAs [002]. The mismatch is about 2.3 % in both directions at room temperature. Such epitaxial relationship results in the following out of plane alignment: SrTiO 3 [110] // GaAs [1-10]. It should be noted that such interface would lead to an alignment of GaAs [111] with SrTiO 3 [1-12] as observed in figure 4.b. 9

10 Figure S7: Ball and stick model showing the GaAs(1-10) plane and the SrTiO 3 (110) plane An amorphous interface can be formed after the oxide growth as reported in several studies [11]. However, for several NWs an abrupt interface was found (Figure S8.a-c). Figure S8: TEM images showing abrupt GaAs / SrTiO 3 interfaces. SrTiO 3 was grown using the two-step method. 10

11 Several NWs exhibit polycrystalline agglomerated SrTiO 3 grains, particularly at the NWs tip as shown in figure S9. Figure S9: TEM image of a NWs tip. SrTiO 3 was grown using the two-step method. It should be noted that the GaAs (1-10) surface is not polar while the SrTiO 3 (110) surface is polar which could be of importance, since the commonly reported interface between GaAs and SrTiO 3 involves the polar GaAs (001) surface and the non-polar SrTiO 3 (001) surface. XPS analysis of SrTiO 3 on GaAs (110) substrate should be performed in order to understand how the SrTiO 3 bonds to the GaAs surface using the two-step growth procedure. 4. X-ray photoelectron spectroscopy analysis and thermal stability: Table S1 shows the XPS fitting parameters of figure 5. The fitting was achieved with Fityk [18]. 11

12 Core level Binding energy Spin orbit gwidth (ev) Voigt parameter (ev) splitting (ev) Ti 2p 3/ Sr 3d 5/2 A Sr 3d 5/2 B Sr 3d 5/2 C Ga 3d 5/2 A Ga 3d 5/2 B As 3d 5/2 A As 3d 5/2 B O 1s A O 1s B O 1s C O 2s & Sr4p Table S1: Fitting parameters used for the peak deconvolution of XPS spectrum shown in Figure 5. The thermal stability of the heterostructure was probed by measuring the Ga 3d, Ti 3p and As 3d core levels during annealing inside the XPS analysis chamber. Figure S10 shows the 12

13 evolution of the XPS core levels during the annealing from room temperature up to 550 C. The sample appears to be stable in UHV for temperatures below 500 C. At 500 C the As 3d intensity of As oxides abruptly decreases while a new component appears at around 41 ev related to GaAs. No significant evolution of the Ga oxide component is observed, which indicates a better stability of the Ga oxides compared to the As oxides [19]. The sample was then cooled down to room temperature and the Ga 3d, As 3d, Sr 3d and Ti 2p core levels were measured again. Interestingly, except for the As oxide desorption (O 1s and As 3d core levels) no other chemical modifications were observed (Figure S11). These results demonstrate the good thermal stability of GaAs core / SrTiO 3 shell NWs and the possibility to desorb the As oxide by a simple annealing. Figure S10: XPS As 3d, Ti 3p and Ga 3d core levels of GaAs / SrTiO 3 NWs measured as a function of the substrate temperature during annealing and cooling. The maximum temperature corresponds to the dashed line. The photon energy was 750 ev. 13

14 Figure S11: XPS core levels of GaAs / SrTiO 3 NWs measured at room temperature. The photon energy was 750 ev. The black (red) curves correspond to the core levels measured before (after) the annealing 5. PL measurements 14

15 Figure S12: PL spectra of the GaAs-based core shell NWs. GaAs NWs without SrTiO 3 (a), GaAs NWs with SrTiO 3 grown at 350 C (b), GaAs NWs with SrTiO 3 grown at 620 C (c), GaAs NWs with SrTiO 3 grown by the two-step method (d) Figure S12 shows the PL measurements performed on GaAs / SrTiO 3 NWs prepared with various growth conditions. After SrTiO 3 growth, the PL of GaAs appears to be systematically deteriorated (Figure S12.b-d) compared to the PL measured before the oxide growth (Figure S12.a) 15

16 Figure S13: Time resolved PL measurements for GaAs / AlGaAs NWs (a) and GaAs / AlGaAs / SrTiO 3 NWs. The lifetime was extracted from monoexponential decay fitting (red line). Measurements were performed at 300K. The time-resolved measurements were performed using a 200 fs pulsed laser emitting at 515 nm with a 50 MHz repetition rate (pulse excitation density of 2.4 µj/cm²). The NW PL signal was analyzed by a spectrometer and a synchronized streak camera with a temporal resolution of 20 ps. Supplementary references: [1] Sibirev N V, Tchernycheva M, Timofeeva M A, et al. Journal of Applied Physics 111, (2012) [2] Sartel C, Dheeraj D L, Jabeen F, et al. Journal of Crystal Growth (2010) [3] Czaban J A, Thompson D A, LaPierre R R. Nano letters 9, 148 (2009) [4] Zhou H L, Hoang T B, Dheeraj D L, et al. Nanotechnology 20, (2009) [5] Armitage R, Tsubaki K. Nanotechnology 21, (2010) 16

17 [6] Kelrich A, Calahorra Y, Greenberg Y, et al. Nanotechnology 24, (2013) [7] Teki R, Datta M K, Krishnan R, et al. Small 5, 2236 ( 2009) [8] Tourbot G, Bougerol C, Grenier A, et al. Nanotechnology 22, (2011) [9] Popovitz-Biro R, Kretinin A, Von Huth P, et al. Crystal Growth & Design 11, 3858 (2011) [10] Cao B Q, Zúñiga-Pérez J, Boukos N, et al. Nanotechnology 20, (2009) [11] G. Delhaye, C. Merckling, M. El-Kazzi, G. Saint-Girons, M. Gendry, Y. Robach, G. Hollinger, Journal of Applied Physics 100, (2006) [12] G. Delhaye, M. El Kazzi, M. Gendry, G. Hollinger, Y. Robach, Thin Solid Films 515, 6332 (2007) [13] G. Niu, G. Saint-Girons, B. Vilquin, G. Delhaye, J.-L. Maurice, C. Botella, Y. Robach, G. Hollinger, Applied Physics Letters 95, (2009) [14] L. Louahadj, R. Bachelet, P. Regreny, L. Largeau, C. Dubourdieu, G. Saint-Girons, Thin Solid Films 563, 2 (2014) [15] B. Meunier, R. Bachelet, G. Grenet, C. Botella, P. Regreny, J. Penuelas, G. Saint-Girons, Journal of Crystal Growth 433, 139 (2016) [16] J. Hao, J. Gao, Z. Wang, D. P. Yu, Applied Physics Letters 87, (2005) [17] J. Hao, J. Gao, H. K. Wong, Appl. Phys. A. 81, 1233 (2005) [18] M. Wojdyr, Journal of Applied Crystallography 43, 1126 (2010) [19] T. Van Buuren, M. K. Weilmeier, I. Athwal, K. M. Colbow, J. A. Mackenzie, T. Tiedjie, Applied Physics Letters 59, 464 (1991) 17