Mater. Res. Soc. Symp. Proc. Vol. 1395 © 2012 Materials Research SocietyDOI: 10.1557/opl.2012.445

Effects of Crystallographic Planes on Focused Ion Beam Milled Patterns of Single Crystal Diamonds

Rustin Golnabi1, Won I. Lee1, Deok-Yang Kim1, and Glen R. Kowach2 1 Bergen County Academies, 200 Hackensack Avenue, Hackensack, NJ 07601, U.S.A. 2 Department of Chemistry, The City College of New York, 160 Convent Avenue, New York, NY 10031, U.S.A.

ABSTRACT

Focused ion beam (FIB) milling of diamonds has been investigated in various ways to create

desired structures on diamonds, but not much research has been reported on the effects of crystal orientation, i.e. {100}, {110} and {111} of diamonds on FIB milling. In our previous work, it was noted that focused ion beam milling may develop preferred etched directions related to the crystal orientation of crystalline diamonds. In order to further investigate the phenomenon, a focused beam of 30 kV Ga+ ions was utilized to generate various patterns on different crystallographic planes of single crystalline diamonds. The morphology of milled patterns has been monitored with various ion currents to find the relationship between crystal orientations of diamonds and their impacts on FIB milled patterns. The work showed significant differences in deformation among different crystal orientations of the single crystal diamond, and the largest area of milling in {111} crystallographic planes.

INTRODUCTION

Diamond has been recognized has an increasingly advantageous material for applications

such as spintronics [1], quantum information processing [2], and bio-sensing [3]. However, because of the difficulty of use provided by diamond’s natural hardness, microfabrication of diamond to precision has become a major focus of recent research. Due to the level of precision it provides, Focused ion beam (FIB) etching is one of the few methods by which diamond can be milled efficiently [4], and it has become a useful option in a variety of nanometer scale fabrication applications, such as circuit editing, microstructural analysis, or nanomachining [5]. This study addresses the use of gallium ion (Ga+) milling, which has also been used in ion beam-induced deposition, of diamond [6]. However due to the mechanical hardness of diamond, a discussion its ion beam milling must include the fabrication challenges associate with it [7]. In a previous study, the re-deposition and nano-terracing effects from FIB results were investigated using a single crystal {110} diamond.

When milling diamond, the crystallographic plane surface which the FIB penetrates can have significant effect on the outcome of the pattern [8]. Depending on the orientation of the diamond, the molecular and atomic structure of the diamond can affect the amount of area and deformation and overall clean FIB milling of the diamond surface.

It has been shown that chemical etching of diamond, particularly Focused ion beam in XeF2 milling, produces high etch rates, particularly high in the {100} face, and low in the {111} face [8]. In other research from Jin et al., however, etching by molten cerium of single crystal diamond substrates showed opposite results, where {111} diamond planes had the highest etch rates and deformation [9]. In this study, varying conditions are used to observe the effects of

surface crystallography of diamond and determine the optimal plane orientation in direct, non-gas-assisted FIB milling procedures.

EXPERIMENTS Imaging and analysis of single-crystal {110), {111}, and {100} orientations diamond was

completed with a FEI Quanta 3D 200i dual beam system which uses Si as the material file. Patterning was done on three high pressure-high temperature synthesized, yellow-colored diamonds. (Figure 1) Each diamond was oriented using carbon tape onto a stage mount with {111},{110}, and {100} planes facing upwards. The diamonds were gold-coated with a thickness of about 3 nm to prevent charging issues before being milled under vacuum. The ion beam source was 30 kV Ga+ with various ion beam currents as described. Default settings of the beam, which were unchanged for these experiments, involved a milling direction starting from the bottom of the patterns to the top (Figure 2).

Figure 1. SEM Micrographs of gold-coated HPHT diamond substrates. (a) {111} orientation. (b) {110} orientation. (c) {100} orientation. All Scale bars represent 500 μm. RESULT AND DISCUSSION SEM-Observed Defects before FIB study

Pits (Figure 3) were observed using scanning electron microscope (SEM) on the surface of

two of crystallographic planes of high pressure-high temperature synthesized diamonds. In Figure 3a, triangular shaped pits were seen on the surface of a {111} plane of a diamond, and likewise in Figure 3b, square shaped ones on the surface of a {100} plane. These are cited frequently as growth defects [10]. The shapes of the defect pits coincide with each plane or atomic arrangement: square shape for {100} and triangular shape for {111}crystallographic plane.

Ion-milled Patterns on Each Crystallographic Plane

In this experiment, five squares with side lengths of 2 μm by 2 μm (square), and a depth of

20 μm were milled into the three {110}, {100}, and {111} diamond planes using a FIB (Figure 2a-c). These patterns were milled to compare with our previous work done on a polished {110} crystalline diamond, at varying currents: 1, 3, 5, 7, and 20 nA. The time required for milling decreased inversely with increasing current.

a b c

Figure 2. Scanning electron micrograph of square patterns milled at various ion beam currents (from left to right 1, 3, 5, 7, and 20 nA) on different crystallographic planes (a) {100}, (b) {110}, and (c) {111}. Scale bars represent 5 μm in length. The shift of 1 nA milled pattern on {100} may have been caused by charging and discarded from data.

a

b

c

Figure 3. Scanning electron micrograph of triangular and square etch pits observed on the surface of the (a) {111} crystal orientation, and the (b) {100} orientation. Stages were tilted to +10° for better image quality. Scale bars represent 1 μm in length.

In our previous work, it was shown that with increased current, there were increased nano-terracing and re-deposition effects showing on the patterns. It was concluded that FIB current could only be increased at the cost of precision of the pattern. In this study, the areas of milled bottom surfaces of the trenches were estimated as well as the areas of the top surface pattern.

While the set parameters for pattern design were identical for each substrate, the actual patterns created by milling varied greatly within each plane as well as among the three planes. As the current was increased, the surface areas for both the top and bottom of the milled pattern increased as shown in Figure 4. While the bottom surface area shows much smaller value compared to the expected 4 μm2(Figure 4 (a)), the top areas are much greater as shown in the Figure 4 (b). The closest area value to the desired pattern were observed on the top surfaces on {100} and {110} planes with lowest current setting of 1 nA as well as the bottom surface on {110}crystallographic plane with highest current setting of 20 nA.

Milling Pattern Accuracy on Different Crystallographic Planes

In order to estimate the milling pattern accuracy, the difference between the top and bottom

surface area was plotted in Figure 5. In the best case scenario, there should not be any difference between the milled top and bottom surface area. The size of pattern used in this research was 2 × 2 × 20 μm3. The aspect ratio of width to the depth is 10. Therefore, it is expected to reveal large difference because of difficulty in the physical removal of materials as the milling process goes on longer. The main purpose of the high aspect ratio was to observe any difference in the milling on different crystallographic planes. Upon plotting the difference between two surface areas as a function of ion currents, it was revealed that the data on {110} and {100} planes showed similar pattern inaccuracy while the inaccuracy of the specific pattern on {111} plane was significantly higher, nearly twice at the lowest ion current milling. The discrepancy between {111} and other planes were smaller at high ion currents.

The depth and milled volume estimation was attempted both changing height of the sample to focus in the bottom of the sample in the electron microscope or 3-D construction of milled

b a

patterns with stereographic imaging, but produced inconclusive results. Therefore, it is not possible to report the milling yield dependence on crystallographic planes.

Figure 4. The (a) bottom surface areas of the squares were measured and plotted as a function of ion current, and the (b) outer top areas of the squares were measured and plotted as a function of ion current. Data for the 1 nA square on the {100} surface was not included because of unanticipated charging issues and deformation in the final pattern.

Figure 5. The difference of top and bottom milled area plotted as a function of ion current. Data for the 1 nA square on the {100} surface was not included because of unanticipated charging issues and deformation in the final pattern.

a b

CONCLUSIONS Through the Ga+ 30 kV focused ion beam milling of varying crystallographic planes of

single crystal diamonds, analyses of generated patterns were able to yield several conclusions and results regarding the dependence of etch rate and deformity on crystallographic plane orientation. As the current increases, the amount of deformity (i.e. graphitization, sidewall terracing) increases and precision or adherence to intended parameters decreases. This results in a much less accurate design, and it has been shown that this inaccuracy is most prevalent in the {111} diamond plane, with nearly twice the amount of deformity in currents 1 through 7 nA. At the highest current (20 nA), there is so much bombardment of the crystal that the orientation is less of an issue. ACKNOWLEDGMENTS

The authors would like to thank Dr. Howard Lerner, Superintendent of the Bergen County

Technical Schools, Russell Davis, principal of the Bergen County Academies, and Edmund Hayward, Technology Director, for their continuous support of chemistry/nanotechnology research program. Furthermore, we would also like to thank Alyssa Calabro, Craig Queenan, and David Becker for their assistance and technical support in the Nano-Structural Imaging Lab at Bergen County Academies.

REFERENCES [1] R. Hanson, O. Gywat, and D. D. Awschalom, Physical Review B, 74, 1203 (2006) [2] W. Chen, P. Chen, A. Madhukar, R. Viswanathan, J. So, MRS Proceedings, 279, 599 (1992) [3] A. Härtl, E. Schmich, J. A. Garrido, J. Hernando, S. C. R. Catharino, S. Walter, P. Feulner, A. Kromka, D. Steinmüller, M. Stutzmann, Nature Materials, 3, 736 (2004) [4] A. Stanishevsky, Thin Solid Films, 398, 560 (2001) [5] S. Reyntjens and R. Puers, Journal of Micromechanics and Microengineering, 11, 287 (2001) [6] D. P. Adams, M. J. Vasile, G. Benavides, and A. N. Campbell, Journal of the International Societies for Precision Engineering and Nanotechnology, 25, 107 (2001) [7] A. A. Tseng, Journal of Micromechanics and Microengineering, 14, R15 (2004) [8] J. Taniguchi, N. Ohno, S. Takeda, I. Miyamoto, M. Komuro, J. Vac. Sci. Technol. B.,16, 2506, (1998) [9] S. Jin, W. Zhu, T. Siegrist, T. H. Tiefel, G. W. Kammlott, J. E. Graebner, M. McCormack, Appl. Phys. Lett., 65, 21 (1994) [10] M. Omar, N. S. Pandya, S. Tolansky, Proc. R. Soc. Lond. A, 225, 1160 (1554) [11] E.S. Sadki, S. Ooi, and K. Hirata, Appl. Phys. Lett., 85, 6206 (2004) [12] M. W. Geis, J. C. Twichell, J. Macaulay, and K. Okano, Appl. Phys. Lett., 67, 1328 (1995)