FREE ELECTRONIC LIBRARY - Dissertations, online materials

Pages:   || 2 | 3 |

«Step Formation on Hydrogen-etched 6H-SiC{0001} Surfaces S. Nie, C. D. Lee§, and R.M. Feenstra* Dept. Physics, Carnegie Mellon University, ...»

-- [ Page 1 ] --

Published in Surf. Sci. 602, 2936 (2008).

Step Formation on Hydrogen-etched 6H-SiC{0001} Surfaces

S. Nie, C. D. Lee§, and R.M. Feenstra*

Dept. Physics, Carnegie Mellon University, Pittsburgh, PA 15213, USA

Y. Ke, R.P. Devaty and W.J. Choyke

Dept. Physics and Astronomy, University of Pittsburgh, Pittsburgh, PA 15260, USA

C. K. Inoki and T. S. Kuan

Dept. Physics, University at Albany, SUNY, Albany, NY 12222, USA

Gong Gu

Sarnoff Corporation, CN5300, Princeton, NJ 08543, USA

The formation of step bunches and/or facets on hydrogen-etched 6H-SiC (0001) and ( 000 1 ) surfaces has been studied, using both nominally on-axis and intentionally miscut (i.e. vicinal) substrates. It is found that small miscuts on the (0001) surface produce full unit-cell high steps, while half unit-cell high steps are observed on the ( 000 1 ) surface.

The observed step normal direction is found to be 1 1 00 for both surfaces. Hence, for intentionally miscut material, a miscut oriented towards this direction produces much better order in the step array compared to a miscut oriented towards a 11 20 direction.

For (0001) vicinal surfaces that are miscut towards the 1 1 00 direction, the formation of surface ripples is observed for 3° miscut and the development of small facets (nanofacets) is found for higher miscut angles. Much less faceting is observed on miscut (000 1 ) surfaces. Additionally, the (0001) surface is found to have a much larger spatial anisotropy in step energies than the ( 000 1 ) surface.

I. Introduction Vicinal surfaces, that is, surfaces cut on some small angle relative to a low-index face, have been widely used in semiconductor epitaxy as templates on which to grow well- ordered overlayers. For the case of SiC, a given polytype (i.e. a particular stacking arrangement) in the substrate can be preserved in homoepitaxial growth by the use of a stepped surface on the substrate [1]. For heteroepitaxy of GaN on SiC, vicinal surfaces of 6H-SiC(0001) have been found to be beneficial for reducing stacking fault density in the GaN [2]. Additionally, self-ordering of steps on semiconductors has attracted attention lately because of its potential use in patterning and fabrication of nanometer scale device structures [3].

There have been many reports concerning surface cleaning and smoothing of SiC to remove polishing damage on as-received substrates. High temperature etching in gaseous environments of H2 [2, 4, 5, 6, 7, 8] or mixed H2 and HCl [9, 10, 11] removes surface scratches effectively and generates atomically smooth surfaces. Step bunching and ordered step-terrace arrays are observed on etched surfaces and have been explained on the basis of differing etch rates for various types of steps [7, 8, 9] or by invoking a model of elastic interactions between steps [11].

§ permanent address: Raytheon RF Components, Andover, MA 01810, USA * author to contact: feenstra@cmu.edu 1 In this paper we report on the formation of step bunches and/or facets on H-etched SiC (0001) and (000 1 ) surfaces. We refer to the former surface as the Si-face and the latter as the C-face. We have studied both nominally on-axis and vicinal surfaces, with miscut angles of up to 12°. We find in general for very small miscut angle (1º) that full unit-cell height steps occur on the Si-face, while for the C-face half unit-cell height steps are observed. Since the observed step normal direction is 1 1 00 (for both the Si-face and the C-face), it is found that a surface miscut in this direction shows much better order in the step array compared to a miscut in the 11 20 direction. For vicinal surfaces that are miscut towards the 1 1 00 direction, faceting is observed on Si-face surfaces whereas the C-face shows almost no faceting. The Si-face results extend over a much larger angular range and show significantly different spatial periods than in previous work [7,8], whereas for the H-etched miscut C-face there are no prior reports that we are aware of.

The differences that we observe between the Si-face and C-face, full unit-cell steps and faceting in the former but not the latter, are interpreted in terms of differences in the etch rates of the various types of steps on the two faces.

II Experimental SiC (0001) Si-face and ( 000 1 ) C-face surfaces were studied, using both 6H and 4H polytype substrates. Results from the two polytypes were similar, and only data from the former are reported here. The 6H substrates were n-type with resistivities in the range 0.03 – 3 ohm-cm. The surfaces were miscut away from the nominal (0001) or ( 000 1 ) planes by angles as large as 12°, with the direction of miscut being towards either the 1 1 00 or 11 20 directions. On-axis substrates as well as substrates miscut at 3.5° towards 11 20 were purchased from Cree Corp. Other substrates used were cut from boule material and mechanically polished with diamond paste with decreasing grit size down to 0.25 μm, and for these samples the miscut angle was checked with x-ray diffraction; these miscut angles have an accuracy of ±0.2°.

A homebuilt strip-heater employing a Ta strip is used for H-etching [12]. The 1 cm × 1 cm sample rests on a 0.001 inch thick Ta strip through which 70 A passes, heating the sample to high temperature. Hydrogen-etching is done in 1 atm of pre-purified (99.995%) hydrogen gas at a flow rate of 11 liters-per-minute for 100 seconds, with the sample held at a temperature in the range 1600 - 1650°C. The hydrogen flow over the SiC starts prior to the rise in temperature and ends after the wafer is cool. Cooling of the sample is done by switching off the power supply, producing a cooling rate of 100°C/s, although cooling rates of ≈10°C/s were also investigated and they produced no change in the results. Temperature is monitored by a disappearing filament pyrometer, with the sample viewed through a window in the etching chamber. Auger electron spectroscopy performed on the samples following the H-etching reveals a very small Ta peak, near the noise limit of the spectrum, implying a Ta surface coverage of about 0.03 monolayers.

After H-etching the substrates are cleaned in acetone, methanol, and deionized water, using an ultrasonic bath. Atomic force microscopy (AFM) is done using either a Park Scientific instrument in contact mode, or a Digital Instruments Nanoscope III in tapping mode. Transmission electron microscopy (TEM) was performed using a JEOL JEM-2010 transmission electron microscope operating at 200 kV.

2 III. Results A. Nominally on-axis surfaces For nominally on-axis substrates, H-etching produces uniformly distributed step-terrace arrays on both Si- and C-face surfaces, as shown in Fig. 1. The step arrays form because of the unintentional miscut of the surface (the average miscut values for the substrates of Figs. 1(a) and (b) are 0.28 and 0.18°, respectively, although these values differ from the local miscut of the observed step arrays of 0.41 and 0.11°, respectively, due to the formation of large-scale low-angle facets on such surfaces as previously described [5]).

Step edges perpendicular to 1 1 00 directions are observed for both faces. Following other authors, we conclude that these step directions have the lowest surface energy among all directions in the (0001) basal plane, or in other words, the H-etching is slowest in these directions [7, 8, 9]. The observed step height from Fig. 1 is 1.5 nm (i.e. six bilayers) for the 6H-SiC Si-face while it is 0.75 nm (three bilayers) for the 6H-SiC Cface, that is, a full unit-cell height in the former case and a half unit-cell height for the latter. The reason for this difference in step height between the Si- and C-faces will be discussed in Section IV.


B. Vicinal surfaces miscut towards 1 1 00 On vicinal surfaces, the width of terraces and the morphology of steps depend on miscut direction and angle. We first consider miscut angles towards 1 1 00, i.e. the same direction as the normal direction for low-energy steps. For small miscuts of 1° the steps maintain their equidistant arrangement. However, at higher miscuts a new phenomenon occurs, as shown in Fig. 2 for 6H-SiC Si-face surfaces miscut at 1° and 3°. Individual steps are clearly resolved on both surfaces, Figs. 2(b) and (d), and they have full unit-cell height.

However, an additional feature seen in both images is a rippling of the surface morphology.

This rippling persists over the entire surface, and has an average period in Fig. 2(a) of about 2 μm and in Fig. 2(c) of about 500 nm. [The line scan in Fig. 2(c) shows only the long-range modulation in surface height, with the individual unit-cell high steps being too closely-spaced to be resolved, although they are clearly seen in Fig. 2(d).] These relatively long-wavelength oscillations are consistent with our prior study in which we found a period of about 300 nm on a surface miscut at 3.5° towards 1 1 00 (although individual atomic steps were not resolved in that work) [2]. In general, this rippling is indicative of step-step separations that are no longer constant, but have some small periodic variation over the surface. Steps are closer spaced at the ascendant part of a ripple than at the descendant part.

It should be noted that this type of arrangement occurs without forming any specific highindex crystal planes (i.e. facets) on the surface. An additional feature evident in Fig. 2(b) is that orientation of the steps is not perfectly constant (i.e. the steps are slightly curved), an issue that we return to in Section V.

With an even higher miscut of 5º the periodic undulations on the surface become much smaller in spatial period, as illustrated in Fig. 2(e). Individual steps of full unit-cell height (1.5 nm) are clearly visible in this image (about 35 steps are apparent, with the step edges extending vertically across the image). These individual steps tend to group together producing undulations with period of 100 nm. The undulation pattern is somewhat irregular on the surface, but nevertheless on the right-hand side of Fig. 2(e) one can clearly see groups of about 6 unit-cell high steps giving rise to an undulation 3 period of about 80 nm. Following prior work we denote these features as nanofacets [8].

The line scan shown in the inset of Fig. 2(e) clearly shows small (0001) terraces separated by the nanofacets.

Nanofaceting of the Si-face persists at higher miscut angles. We show in Fig. 3(a) an AFM image of the Si-face with a 12° miscut, and Fig. 3(b) shows data for the C-face with the same miscut. In both cases a closely-spaced array of steps is visible on the images, with the apparent step separation being significantly less for the C-face than for the Siface. We find an average separation between linear (approximately periodic) features seen in the images of 19 nm for the Si-face and 7 nm for the C-face. For comparison, given the 12° miscut of the surfaces, if the surface consists of unit-cell high steps (as occurs for the Si-face with low miscut angle) then we would expect a 7.0 nm step-step separation, or if the surface consists of half unit-cell high steps (as occurs for the C-face with low miscut angle) then we would expect a 3.5 nm separation. Thus, we conclude from the data of Fig. 3 that for the Si-face we appear to be resolving step bunches of typical height 2 – 3 unit cells, whereas for the C-face we are resolving steps of single unit-cell height.

Our conclusions based on the AFM data are supported, and clarified, by crosssectional TEM of the same samples, as shown in Fig. 4. On the Si-face, Fig. 4(a), faceting of the surface is clearly observed, with the appearance of ( 1 1 0 n ) facets having n ≈ 12 and making an angle of ≈25° to the (0001) planes. A similar facet angle for the Si-face has been reported previously [8]. From the TEM data we find typically 4 – 5 unit cell heights separating the observed (0001) terraces, which is similar (slightly larger) than the value deduced from the AFM images. For the C-face, minimal faceting is observed; any apparent step bunches involve only single or perhaps double unit-cell height steps. We cannot distinguish on the TEM image of Fig. 4(b) any distinction between single or half unit-cell high steps, but in any case the image is consistent with the AFM data of Fig.

3(b) that suggests single unit-cell high steps for this highly miscut C-face surface with no evidence of nanofaceting.

C. Vicinal surfaces miscut towards 11 20 We now turn to surfaces miscut towards 11 20, a miscut direction that is commonly used in epitaxial growth [13]. Since this direction does not correspond to the low-energy step direction for our H-etched surfaces, we find in general much less order in the resulting step array than for the 1 1 00 miscut direction of the previous section.

Results for the Si-face are shown in Figs. 5(a) and (b). At 3.5° miscut we observe zigzag step edges with half unit-cell height (0.75 nm), as shown in Fig. 5(a). The average direction of each step has step normal towards the 11 20 miscut direction, but locally the step edges appear to have a zigzag shape alternating portions being locally perpendicular to 1 1 00 directions. We thus interpret the step edge morphology as having many kinks, in the manner described by Nakajima et al. in Fig. 6 of their work [8]. It should be noted that this morphology is dependent on the etching method and conditions. Other etching methods (HCl+H2) have given straight step edges with 1120 step normal, or “triangular depressions”, depending on the etching temperature [10]. The occurrence in our study of the local 1 1 00 step morphology is further evidence that, for H-etching, the etching rate along 1 1 00 is slower than that along 11 20. At a 4 higher miscut angle of 12°, no average step orientation is found, as shown in Fig. 5(b).

The steps are seen to meander all over the surface. Saitoh et al. [14] have reported similar surface morphology of an epilayer on a 4H-SiC Si-face substrate with 18º miscut towards

11 20. The reason for this change in morphology as the miscut angle increases is not well understood at present, but it is perhaps not so surprising that for a surface with such a high kink density (i.e. as occurs for steps with average orientation of 11 20 ) a nonuniform distribution of step spacing results.

Pages:   || 2 | 3 |

Similar works:

«The BMB Weekly Vol. 41, No. 40, September 29-October 3, 2008 nd Please send submissions to: Katie Gallagher at galla134@msu.edu or 105B Biochemistry (mailbox on 2 floor) Calendar Monday, September 29 Department of Physiology Cancer Faculty Candidate Seminar: Explore novel epigenetic targets for breast cancer therapy. Yi Huang, Johns Hopkins University, 9:00 a.m., 1425 Biomedical and Physical Sciences. Chemistry Seminar: Presentation by Kapil Lokare, Michigan State University, 11:20 a.m., 136...»

«Uncloaking globular clusters in the inner Galaxy1 Javier Alonso-Garc´ ıa Departamento de Astronom´ y Astrof´ ıa ısica, Pontificia Universidad Cat´lica de Chile, o 782-0436 Macul, Santiago, Chile Department of Astronomy, University of Michigan, Ann Arbor, MI 48109-1090 jalonso@astro.puc.cl Mario Mateo Department of Astronomy, University of Michigan, Ann Arbor, MI 48109-1090 mmateo@umich.edu Bodhisattva Sen Department of Statistics, Columbia University, New York, NY 10027...»

«L-AP “Exploring Creation With Astronomy” Lessons 1-14 Lapbook Package PLEASE NOTE: This product includes BOTH lapbooks for this book. One lapbook covers lessons 1-6, and the other covers lessons 7-14. This lapbook has been specifically designed for use with the book, “Exploring Creation with Astronomy” by Jeannie Fulbright and Apologia Science. Templates are printed with colors that best improve information retention according to scientific research. Designed by Cyndi Kinney of...»

«Measuring the Atmospheric Influence on Differential Astrometry: a Simple Method Applied to Wide Field CCD Frames N. Zacharias1 U.S. Naval Observatory, 3450 Mass. Ave. N.W., Washington D.C. 20392, Electronic mail: nz@pyxis.usno.navy.mil Received ; accepted PASP revised manuscript, 12 Aug 96, proofs to NZ 1 with Universities Space Research Association (USRA), Division of Astronomy and Space Physics, Washington D.C., based on observations made at KPNO and CTIO –2– ABSTRACT Sets of short...»

«John Redden Copyright 2008 © Cosmos Heaven Awakening I have no beginning or. and. Click. The grains of dirt on my plain don't live. A plain I know for all conscious existence spreads before me. The colors change slightly to make up time periods. This awakening period is unusual. There was a clear impression Ideea had on the borderlands of her consciousness. She had visions of a brown haired woman in a cool and comfortable place very different from the omnipresent Waste. Then the words that...»

«Formation of the Galactic Halo. Inside and Out ASP Conference Series, Vol. 92, 1996 Heather Morrison and Ata Sarajedini, eds. Globular Clusters as Fossils of Galaxy Formation Richard B. Larson Yale Astronomy Department, Box 208101, New Haven, CT 06520-8101 Abstract. The globular clusters in the halos of large galaxies like our own are almost certainly fossil remnants of the early star-forming subsystems from which these galaxies were built. The ages of the halo clusters in our Galaxy indicate...»

«DOI: 10.5433/1984-3356.2015v8n15p313 Una forma territorial alternativa: la tribu de Coliqueo en la pampa bonaerense An alternative territorial form: indigenous tribe of Coliqueo in the Buenos Aires pampas Melina Yuln Graciela Silvestri RESUMEN Durante gran parte del siglo XIX la pampa de Buenos Aires fue dominio de los indígenas, propiciando así el retrato de un espacio salvaje. Luego de ser controlado por el Estado este territorio sufrió un proceso de transformación en el cual se...»

«El currículo de educación infantil como un tejido de experiencias, relaciones y saberes Marisol Anguita y Fernando Hernández' Desde los márgenes del currículo Consideramos que 'currículo' hace referencia a la relación que una maestra construye con cada uno de los niños y las niñas de su grupo. No es un listado de objetivos, competencias y contenidos. Es, por encima de todo, una experiencia de relación que posibilita interrogar-nos, descubrir y dar sentido a nuestra relación con el...»

«The Shapes of Dense Cores and Bok Globules Barbara S. Ryden 1 Department of Astronomy, The Ohio State University, 174 W. 18th Ave., Columbus, OH 43210 ABSTRACT The shapes of isolated Bok globules and embedded dense cores of molecular clouds are analyzed using a nonparametric kernel method, using the alternate hypotheses that they are randomly oriented prolate objects or that they are randomly oriented oblate objects. In all cases, the prolate hypothesis gives a better fit to the data. If Bok...»

«Astronomy Cast Episode 76: Lagrange Points Fraser Cain: Gravity is always pulling you down, but there are places in the solar system where gravity balances out. These are called Lagrange points and space agencies use them as stable places to put spacecraft. Nature is on to them and has already been using them for billions of years. Before we get on to it, let’s talk about pronunciation. I said it Lagr-ahunge points. Is that okay? Dr. Pamela Gay: I have heard it said Lagr-ahunge points,...»

<<  HOME   |    CONTACTS
2016 www.dissertation.xlibx.info - Dissertations, online materials

Materials of this site are available for review, all rights belong to their respective owners.
If you do not agree with the fact that your material is placed on this site, please, email us, we will within 1-2 business days delete him.