This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and

Transcription

1 This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier s archiving and manuscript policies are encouraged to visit:

2 Journal of Crystal Growth 310 (2008) Contents lists available at ScienceDirect Journal of Crystal Growth journal homepage: Biaxially oriented CaF 2 films on amorphous substrates H.-F. Li a, T. Parker a, F. Tang a, G.-C. Wang a,, T.-M. Lu a, S. Lee b a Department of Physics, Applied Physics and Astronomy, Rensselaer Polytechnic Institute, 1C25 Science Center, 110, 8th Street, Troy, NY , USA b US Army Armament Research, Development and Engineering Center, Benét Labs, Watervliet, NY 12189, USA article info Article history: Received 10 March 2008 Received in revised form 23 April 2008 Accepted 23 April 2008 Communicated by P. Rudolph Available online 7 May 2008 PACS: JK abstract Biaxially oriented CaF 2 films have been created by first using an oblique incidence vapor flux to deposit CaF 2 onto an amorphous surface to form vertically aligned nanorods which served as seeds to grow a more continuous CaF 2 capping layer under a subsequent normal incidence flux deposition. The entire film possesses a unique {111}/121S biaxial texture as shown by X-ray pole figure analysis and transmission electron microscopy (TEM). This unique texture formation is a result of shadowing and surface diffusion effects. This biaxially oriented film on an amorphous substrate may be useful as a buffer layer to grow active semiconductor devices. & 2008 Elsevier B.V. All rights reserved. Keywords: A1. Crystal morphology A1. Crystal structure A1. Nanostructures A3. Physical vapor deposition process B1. Calcium compounds I. Introduction There has been intense interest in the growth of biaxially textured films on amorphous substrates. The biaxially textured films are not exactly single crystals, but they have strongly preferred crystallographic orientations in both the out-of-plane and in-plane directions [1,2]. Biaxial films have been used as buffer layers for subsequent growth of highly oriented (both the out-of-plane and in-plane) high T c superconducting films to achieve high current density. Techniques such as ion beamassisted deposition (IBAD) [3], rolling-assisted biaxially textured substrate (RABiTS) [4], and oblique angle deposition (or inclinedsubstrate) [5 7], have been adopted to create biaxially textured films. Very recently, it was proposed that a biaxially textured film on glass or other amorphous substrates may be used as a buffer layer to grow high-quality semiconductor films for efficient and low-cost solar cells [8,9] and other applications in optoelectronics. Despite many challenges in materials compatibility, biaxial quality, and surface topology, this methodology is intriguing and if successful the impact could be substantial. Oblique angle deposition is a simple, inexpensive, and scalable technique for creating biaxially oriented films. A good example is Corresponding author. Tel.: ; fax: address: wangg@rpi.edu (G.-C. Wang). oblique angle deposited biaxial MgO films, which have been used as the buffer layer for the growth of oriented high T c superconductor films [5,7]. However, the growth of semiconductor films on these oxide buffer layers could cause an undesirable interface reaction with the semiconductor materials. Semiconductor materials would thus be oxidized when deposited on the MgO surface and the formation of biaxial texture would be disrupted. In the present work, we focus on the creation of biaxial CaF 2 films using the oblique angle deposition technique on Si wafers covered with a layer of amorphous native oxide. CaF 2 has been known as an excellent buffer layer material for the growth of a variety of semiconductor materials including Si [10 14] and compound semiconductors [15 17]. The heteroepitaxial structures with CaF 2 have promising applications in many advanced devices [14], such as resonant tunneling diodes [12] and electrooptical modulators [14,18]. In the past, high-quality CaF 2 buffer layers have only been achieved on single-crystal Si substrates. Here we adopted a two-step process to grow biaxially oriented CaF 2 films on amorphous substrates. Basic mechanisms that lead to this final biaxially textured film are discussed. 2. Experimental procedure The growth of the CaF 2 nanorods was performed at near room temperature (no intentional heating) in a high-vacuum thermal /$ - see front matter & 2008 Elsevier B.V. All rights reserved. doi: /j.jcrysgro

3 H.-F. Li et al. / Journal of Crystal Growth 310 (2008) evaporation system with a base pressure of Torr. The detail of the setup was described elsewhere [19]. The CaF 2 source (Alfa Aesar, 99.9% pure, in the form of flakes) was placed in a tungsten evaporation boat. The distance between the source and the substrate was about 25 cm. Si wafer pieces (1 1cm 2 in dimension) covered with a native oxide were used as the substrate. We grew a series of films with an oblique angle a (angle between the incident flux and the substrate normal) ranging from 01 to 751. The texture of the nanorod films was evaluated through X-ray diffraction and X-ray pole figure analyses. X-ray pole figures were obtained by collecting multiple diffraction patterns using an area detector in a Bruker D8 Discover diffractometer (Cu target, wavelength ¼ nm) [1]. Transmission electron microscopy (TEM, model JEOL 2010, 200 kv) was used to examine how the texture evolves at different stages of growth in more detail. A conventional cross-section TEM sample preparation method was adopted here. Two pieces of sample were glued together face to face using epoxy. The glued sample was subjected to grinding and polishing to a thickness of 20 mm and followed by ion milling to reach electron transparency. The morphology of the nanostructured film was studied by field emission scanning electron microscopy (SEM, Carl Zeiss Supra SEM 1550). 3. Results and discussion In general, the nanorods are tilted. However, when a is around 651, the nanorods are oriented almost vertically. Fig. 1a and b show the SEM side view and top view images of a 750-nm-thick nanorod film with the incident flux coming from the left (a651) indicated by the arrow. The deposition rate was 15 nm/min. These images show that the tip of the rods are faceted and the nanorod film appears to have the similar roof tile structure reported in the literature for MgO films grown by oblique angle deposition [20] CaF 2 Counts (arb. units) θ (degs.) Fig. 1. (a) SEM side view and (b) SEM top view of 750-nm-thick CaF 2 nanorods grown by oblique angle (651) vapor deposition incident from the left. (c) A two-dimensional (2D) X-ray diffraction pattern of the CaF 2 nanorods film. (d) The 2y plot constructed by integrating a narrow slice (711) of 2D diffraction patterns along the dashed line. (e) X-ray pole figure showing the [111] out-of-plane orientation of 750-nm-thick CaF 2 nanorods. (f) A schematic showing the side view of a faceted CaF 2 nanorod with the [111] direction perpendicular to the substrate, the ½1 1 1Š direction facing towards the incident flux, and the [121] direction in the facet plane.

4 3612 H.-F. Li et al. / Journal of Crystal Growth 310 (2008) Fig. 1c is an image of the raw two dimensional (2D) X-ray diffraction pattern of the CaF 2 nanorods film deposited at 651 incidence angle. There is a sharp (111) arc nearly centered about the substrate normal. Fig. 1d isa2y plot that was formed by integrating a narrow slice (711) along the substrate normal (dashed line in Fig. 1c) of the 2D diffraction data. From this plot we can clearly see a dominant (111) out-of-plane texture as compared with the powder diffraction intensities (ICCD card ). The other diffraction peaks from CaF 2 are very weak, but some small features can be observed (for example, a (2 2 0) peak at 2yE471 and a (4 0 0) peak at 2yE681). To reveal the biaxial texture of the film, the (111) pole figure was constructed, as shown in Fig. 1e. CaF 2 is a cubic structure, the adjacent /111S directions are separated by theoretically. In the pole figure, the angular distances between each two concentrated {111} pole positions are close to 701, consistent with the theoretical calculation. The well-defined pole intensity concentration in the pole figure thus indicates that the CaF 2 film is biaxially oriented, i.e., it has not only an out-of-plane preferred orientation but also a clear in-plane preferred orientation. The existence of the center (111) pole indicates that the crystallographic direction of the film normal to the substrate is the [111] direction. It has been argued that the out-of-plane texture orientation is controlled by surface energy [21 23], because the mobility of deposited atoms is high on the plane with the minimum energy and would spread quickly on this plane to dominate the growth. In an oblique angle deposition, the minimum energy plane would face towards the incidence flux and would give a texture tilted towards the flux. For CaF 2, the minimum energy plane is {111} [24]. In the present case, the faceted surface facing the incidence flux is the (11 1) plane. This plane is equivalent to any other planes in the {111} family. Here a specific plane is used to facilitate the discussion. As for the inplane orientation, crystals with the azimuthal orientation having the highest vertical growth rate would survive [22,25] and would define the in-plane texture. A schematic of the growth front of CaF 2 nanorods is shown in Fig. 1f based on our X-ray pole figure analysis and TEM analysis presented later. It is concluded that the fastest vertical growth direction appears to be the /121S direction for CaF 2 deposition. We use {111}/121S to label this biaxial structure. Similar to the MgO biaxial films reported in the Refs. [5,6], the surface of our CaF 2 nanorod film is quite porous as seen in Fig. 1b. The porous surface is not desirable to be used as a buffer layer for further growth of semiconductor films. We then deposited an additional 250-nm-thick CaF 2 as a capping layer on top of the nanorod film by normal incidence deposition. We expect that the nanorods would serve as seeds for the homoepitaxial growth of the capping layer. Due to the atomic steering effect, the structure on top of the nanorods would fan-out during deposition [26] to cover most gaps in between the nanorods. The origin of the atomic steering effect is due to the long-range van der Waals force that tends to attract the incoming atoms which would then stick to the side walls and cause sideways growth. A slower deposition rate of 9 nm/min was used to grow this capping layer. Fig. 2a shows the SEM side view image of the 250 nm capping layer (region II) on top of the 750 nm nanorods film (region I). Fig. 2b shows the top view of the capping layer (region II) which contains three-fold symmetric faceted nanocrystals. The capping layer has a multilayered inverted-pyramid structure. The top surface of the Fig. 2. (a) SEM side view of the 750-nm-thick CaF 2 nanorods (region I) grown by oblique angle deposition and 250-nm-thick CaF 2 capping layer (region II) grown by normal incidence deposition. (b) SEM top view of the 250-nm-thick CaF 2 capping layer. (c) X-ray pole figure showing the [111] out-of-plane orientation of the entire film with 250- nm-thick CaF 2 capping layer on 750-nm-thick CaF 2 nanorods. (d) A schematic showing the top view of one faceted CaF 2 capping layer with an inverted pyramid consisting of three {110} planes.

5 H.-F. Li et al. / Journal of Crystal Growth 310 (2008) film is corrugated but the coverage of the surface is much larger compared with that of the roof tile structure of the nanorod film which contains large gaps and holes, as seen in Fig. 1b. An X-ray pole figure shown in Fig. 2c was taken from the film with the capping layer. The pole figure looks qualitatively similar to that shown in Fig. 1c, which is from the nanorods film without the capping layer. This indicates that the capping layer was grown epitaxially on top of the nanorods and has maintained the same crystallographic orientation as the nanorods film. Because X-ray penetrates deep into the film, the pole figure reflects the average texture information from the entire thickness of the film. The capping layer consists of inverted pyramids. The tip angle is estimated to be from the SEM side view shown in Fig. 2a. This value is close to the tip angle (1251) of the inverted pyramid structure with the three {110} planes as side faces. Fig. 2d is a schematic of the top view of the capping layer showing the three proposed {110} side facets. Also, there is an island nucleated at the center of the inverted pyramid. This island would serve as a seed to grow a subsequent {110} faceted layer. To examine how the texture evolves at different stages of growth in more detail, we performed a TEM analysis of the sample shown in Fig. 2. Fig. 3 is a TEM cross-sectional view of the sample and the corresponding selective area diffraction patterns from four circled regions. The zone axis is close to the /110S direction. It can be seen from the figure that the broken ring diffraction pattern from circle 1 region shrinks to a nearly spot-like diffraction pattern from circle 3 region when the electron beam moved up along the nanorod height (marked as Region I in the figure). This suggests that the biaxial orientation becomes progressively stronger as the nanorods grow taller. In the capping layer (marked Region II ), where the deposition changed into normal incidence deposition, the nearly spot-like diffraction pattern from circle 4 in region II remains unchanged. We tested different locations along the capping layer region, similar diffraction patterns were obtained. This indicates that the capping layer is grown epitaxially on the top of the nanorods. In Fig. 3, the arrows showing the [2 0 2], [111], and ½1 11Š directions in region II are inferred from the TEM diffraction pattern. The extraction of the spread of the in-plane biaxial orientation at the surface of the film is challenging. X-ray pole figure represents the pole concentration from the entire film, not just from the surface. From Fig. 1c, the spread (full-width-halfmaximum) of the in-plane orientation of the 750-nm-thick nanorods film from the phi-scan, is estimated to be 20.91, while the spread in the film with the additional 250-nm-thick capping layer estimated from Fig. 2c is The spread of the pole becomes narrower after the capping layer is deposited. Therefore the biaxial texture of the capping layer is better than that of the nanorod film. However, the exact spread from the surface biaxial texture of the film, which is expected to be even smaller, is still not known. One cannot extract the spread of the surface texture from a diffraction pattern (such as the ones shown in Fig. 3), because the TEM diffraction pattern is obtained from a particular direction. 4. Conclusions In summary we report the creation of a 750-nm-thick biaxially oriented CaF 2 nanorod film on an amorphous substrate using the oblique angle thermal evaporation deposition technique with an incident flux inclined at 651 with respect to the surface normal. The nanorods have a {111}/121S biaxial orientation with a (11 1) facet facing the incident flux. This nanorod surface is used as a seed to create a 250 nm epitaxial capping layer by a second CaF 2 deposition with a normal incidence flux. This unique CaF 2 film with a capping layer may be useful as a buffer layer to grow an active semiconductor film such as CdTe or polycrystalline Si for energy conversion devices in solar cell applications. Acknowledgment This work is supported in part by NSF no TP is supported by Department of Education GAANN fellowship. References Fig. 3. The left panel shows a cross-section bright field TEM image of 750-nmthick CaF 2 nanorods (region I) and 250-nm-thick CaF 2 capping layer (region II) indicated in Fig. 2a. The right panel shows selected area electron diffraction (SAED) patterns corresponding to circled areas numbered 1 3 in region I and circled area 4 in region II. The zone axis is /110S. The out-of-plane direction [111], the faceted ½1 11Š direction facing the incident flux, and the [2 0 2] direction are inferred from the SAED. [1] F. Tang, T. Parker, G.-C. Wang, T.-M. Lu, J. Phys. D: Appl. Phys. 40 (2007) R427. [2] R.T. Brewer, H.A. Atwater, Appl. Phys. Lett. 80 (2002) [3] Y. Iijima, K. Kakimoto, Y. Yamada, T. Izumi, T. Saitoh, Y. Shiohara, MRS Bull. 29 (2004) 564. [4] A. Goyal, M.P. Paranthaman, U. Schoop, MRS Bull. 29 (2004) 552. [5] Y. Xu, C.H. Lei, B. Ma, H. Evans, H. Efstathiadis, M. Rane, M. Massey, U. Balachandran, R. Bhattacharya, Supercond. Sci. Technol. 19 (2006) 835. [6] M.P. Chudzik, R.E. Koritala, L.P. Luo, D.J. Miller, U. Balachandran, C.R. Kannewurf, IEEE Trans. Appl. Superconductivity 11 (2001) [7] B. Ma, M. Li, R.E. Koritala, B.L. Fisher, A.R. Markowitz, R.A. Erck, S.E. Dorris, D.J. Miller, U. Balachandran, IEEE Trans. Appl. Superconductivity 13 (2003) [8] A.T. Findikoglu, W. Choi, V. Matias, T.G. Holesinger, Q.X. Jia, D.E. Peterson, Adv. Mater. 17 (2005) [9] C.W. Teplin, D.S. Ginley, H.M. Branz, J. Non-Cryst. Solids 352 (2006) 984. [10] J. Wollschlager, C. Deiter, M. Bierkandt, A. Gerdes, M. Baumer, C.R. Wang, B.H. Muller, K.R. Hofmann, Surf. Sci. 600 (2006) [11] L.J. Schowalter, R.W. Fathauer, J. Vac. Sci. Technol. A 4 (1986) [12] M.A. Olmstead, in: A.W.K. Liu, M. Santos (Eds.), Thin Films: Heteroepitaxial Systems, World Scientific, Singapore, [13] D.Y. Kim, B.J. Ahn, S.I. Moon, C.Y. Won, J. Yi, Sol. Energy Mater. Sol. Cells 70 (2002) 415. [14] C.R. Wang, B.H. Müller, E. Bugiel, T. Wietler, M. Bierkandt, K.R. Hofmann, P. Zaumseil, J. Vac. Sci. Technol. A 22 (2004) [15] H. Zogg, J. Masek, C. Maissen, S. Blunier, H. Weibel, Thin Solid Films 184 (1990) 247.

6 3614 H.-F. Li et al. / Journal of Crystal Growth 310 (2008) [16] G. Breton, M. Nouaoura, C. Gautier, M. Cambon, S. Charar, M. Averous, V. Ribes, Surf. Sci. 406 (1998) 63. [17] V. Mathet, P. Galtier, F. Nguyen-Van-Dau, G. Padeletti, J. Olivier, J. Crystal Growth 132 (1993) 241. [18] L.J. Schowalter, R.W. Fathauer, CRC Crit. Rev. Solid State Mater. Sci. 15 (1989) 367. [19] A.K. Kar, P. Morrw, X.-T. Tang, T.C. Parker, H. Li, J.-Y. Dai, M. Shima, G.-C. Wang, Nanotechnology 18 (2007) [20] P. Ghekiere, S. Mahieu, G. De Winter, R. De Gryse, D. Depla, Thin Solid Films 493 (2005) 129. [21] For review see H. Huang, in: S. Yip (Ed.), Handbook of Multiscale Materials Modeling, Springer Science and Business Media, 2005, p [22] A. van der, Drift, Philips Res. Rep. 22 (1967) 267. [23] C.G. Johansen, Hanchen Huang, Toh-Ming Lu, Appl. Phys. Lett. 91 (2007) [24] H. Shi, R.I. Eglitis, G. Borstel, Phys. Stat. Sol. (B) 242 (2005) [25] F. Tang, C. Gaire, D.-X. Ye, T. Karabacak, T.-M. Lu, G.-C. Wang, Phys. Rev. B 72 (2005) [26] D.-X. Ye, T.-M. Lu, Phys. Rev. B 75 (2007)