SYMPOSIUM L

11:00 AM *L1-S2.1 (invited)
Nanoscale Morphology Control Using Ion Beams. (#428) Michael J. Aziz, Harvard School of Engineering and Applied Sciences, Harvard University, Cambridge, Massachusetts, USA.

Low energy ion irradiation of a solid surface can be used to control surface morphology on length scales from 1 micron to 1 nanometer. Focused or unfocused ion irradiation induces a spontaneous self-organization of the surface into nanometer-sized ripples, dots, or holes; it also induces diameter increases and decreases in a pre-existing nanopore by a tradeoff between sputter removal of material and stimulated surface mass transport. Experiments will be reviewed that illuminate the kinetics of evolution of the surface morphological instability; the influence of initial and boundary conditions on guiding the self-organization; the development of shock fronts that sharpen features at sufficiently steep angles; and the kinetics governing the fabrication of nanopores for single-biomolecule detectors.

11:30 AM L1-S2.2
Effect of Ion Bombardment on Au Nanoparticles on Top of Various Substrates. (#333) Paul Ziemann, Andreas Klimmer, Johannes Biskupek, Ute Kaiser; Solid State Physics, Universität Ulm, Germany.

Based on advanced preparational techniques, which exploit the self-organization of inverse micelles loaded with metal precursors, highly ordered arrays of metal nanoparticles were prepared on various sufficiently flat substrates. The technique additionally offers the advantage of controlling the size of the nanoparticles as well as their interparticle distance. Au nanoparticles prepared along this micellar route on top of Si/SiO₂ (partly native oxide, partly thermally grown oxides up to 100 nm thickness) and sapphire substrates formed the starting point of systematic ion bombardments. For this purpose, typically 200 keV Ar+ ions were bombarded onto the Au nanoparticle arrays at ambient temperature. Due to the narrow size distribution of the Au particles with diameters ranging between 2.5 nm and 15 nm, the effect of ion irradiation on these nanoparticles could be accurately studied as a function of their size. It turned out that on Si/SiO₂ the Au particles were surprisingly stable conserving their almost spherical shape during the bombardment, which, however, resulted in completely buried particles at medium ion fluences (typically 1015 cm-2). For higher fluences, eventually the Au nanoparticles are forced into metastable solution within the Si substrate. Corresponding experiments using 200 keV Xe+ projectiles indicate that the ion induced sinking of the nanoparticles scales with the average displacements per atom (dpa). It is important to note that the Au nanoparticles on top of sapphire do not collectively sink into the substrate. Rather, the observed ion induced diameter decrease in that case can be attributed to standard sputtering. Details of these processes as revealed by Atomic Force Microscopy (AFM) as well as Transmission Electron Microscopy (TEM) will be given and discussed in terms of capillary forces and surface/interface energies. We would like to thank A. Ethirajan, B. Kurbanjan and M. Ozawa for experimental help during the early stage of this study and to Landesstiftung Baden-Württemberg "Functional Nanostructures" as well as to Deutsche Forschungsgemeinschaft, SFB 569 for financial support.

11:45 AM L1-S2.3
Elongation of Embedded Metallic Nanocrystals under Swift Heavy Ion Irradiation. (#560) Patrick Kluth¹, Raquel Giulian¹, David J. Sprouster¹, Claudia S. Schnohr¹, Leandro L. Araujo¹, Aidan P. Byrne², David J. Cookson, Mark C Ridgway¹; ¹Electronic Materials Engineering Department, Research School of Physical Sciences and Engineering, The Australian National University, Australian Capital Territory, Australia ; ²College of Science, The Australian National University, Australian Capital Territory, Australia.

Metallic nanocrystals (NCs) embedded in SiO[sub 2] show interesting linear and non-linear optical properties with a high potential for technological applications including optical filters, memories and switching devices. Upon swift heavy ion irradiation spherical nanocrystals can change their shape to form aligned, narrow elongated nanorods, thus introducing a strong optical anisotropy in the material. This intriguing property was first observed in 2003 by D'Orleans et al [1], however, the mechanisms behind the transformation are still under debate. We have investigated the shape transformation on Pt, Co and Au NCs as a function of ion energy, ion fluence and NC size. The NCs were fabricated by ion beam synthesis in 2 um thick SiO[sub 2] layers on Si substrates and subsequently irradiated with Au ions at energies between 27 and 185 MeV. The shape transformation was studied using transmission electron microscopy (TEM) and synchrotron-based small-angle x-ray scattering (SAXS). SAXS measurements were performed at the Advanced Photon Source, Chicago, USA. The transformation process is characterized by an energy dependent saturation width of the resulting nanorods and a related minimum NC size for the transformation to occur. Interestingly, the saturation width of the nanorods is largely independent of the NC material, although the studied materials possess very different properties (melting point, surface energy etc.). In fact, it appears that the observed shape change is strongly related to the dimension of latent tracks that form in the SiO[sub 2] at the given irradiation energies. This is confirmed by additional SAXS measurements of the ion tracks in SiO[sub 2] without NCs. Additionally, dissolution of some of the metal into the matrix during irradiation was observed and quantified. Our results indicate that melting of the NC material and subsequent redistribution in the ion track is a likely mechanism involved in the NC shaping process. [1] C. D'Orleans et al, Phys. Rev. B 67 (2003) 220101(R).

12:00 PM *L1-S2.4 (invited)
Novel Surface Nanostructures by Ion Irradiation Through Nanosphere Lithography Masks. (#1113) Jörg K. N. Lindner, Cornelia Seider, Daniel Kraus, Michael Weinl, Bernd Stritzker; Universität Augsburg, Institut für Physik, Augsburg, Germany.

Recently, a range of new techniques has been developed to form novel functional nanostructures at the surface of solids, and the hype in these new techniques seems to supersede the interest in established surface modification methods. Ion implantation is well-known as a materials-general precision tool to change the atomic structure and near surface composition of solids. The aim of this talk is to demonstrate that ion implantation can be most beneficially combined with such new nano-techniques to create novel surface nanostructures. To this end experiments are presented in which we performed ion irradiations of silicon wafers covered with colloidal masks formed by nanosphere lithography (NSL). NSL is an emerging bottom-up technique for creating ordered arrays of equally sized, sub-100 nm features on surfaces. The technique is based on the self-organised arrangement of spheres from a colloidal suspension on a surface mostly in form of a hexagonally close packed monolayer. These layers can be used as a shadow mask with triangular mask openings in between each triple of adjoining spheres. The size and pitch of mask openings are precisely defined by the diameter of colloidal spheres, typically in the range of 100-1000 nm. NSL masks used here were formed from aqueous suspensions of either SiO2 or polystyrene beads. It will be shown that ion irradiation can be used to tailor the shape of mask openings by inducing the formation of necks in between adjacent spheres, an effect which we call ion beam induced sintering (IBSI) and which is supposed to be based on the large local surface tensions acting at the contact points of nanospheres. IBSI is thus expected to occur also in the ion irradiation of other nano-objects. We shall also demonstrate that ion irradiation can be useful to locally mix mask material and the substrate, and to perform localized implantations into the substrate surface.

LUNCH 12:30 PM - 2:00 PM

2:00 PM *L1-S3.1 (invited)
Molecular Ion Implantation for CMOS Processing. (#1254) Dale Jacobson, SemEquip, Inc. Billerica, Massachusetts, USA.

The use of large dopant containing molecules for semiconductor doping and large non-doping molecules for materials modification will be described. Throughput advantages from boron effective beam currents as high as 45 mA will be shown. Beam currents of this magnitude are now used in the production of high performance memory chips. The ion source and beamline designs will be discussed as well as many device performance enhancements that result from large molecule ion implantation. Self-amorphization, increased activation, and elimination of end-of-range damage will all be discussed and examples from CMOS devices presented. The elimination of end-of-range defects can reduce device leakage by up to two orders of magnitude. Increased activation results in devices that can have more than 20% higher drive current. Large carbon containing molecules such as C₁₆H₁₀ can be effectively used for amorphous layer formation without end-of-range damage, Transient Enhanced Diffusion control, and stress engineering. Charge carrier mobility can be enhanced by more than 40% through the use of carbon induced tensile stress in NMOS channels. Sub-50 nm device data indicating up to 47% higher drive current due to 8E15/cm² carbon implants from C₇H₇ molecules will be shown. Annealing techniques applicable to molecular implantation such as laser, flash, solid phase epitaxy, and spike will all be presented. Sheet Rho vs. Junction Depth plots will be utilized to show the effectiveness of annealing. These large molecules allow for low energy ion implantation without the use of deceleration, thus eliminating high energy tails on the beam due to neutrals. Today, with doses as large as 5E16/cm² for Dual Poly Gate doping, energy contamination as low as 1:10⁴ can have a very negative impact on device performance. In addition to energy contamination-free beams, this technology results in ion beams with very high angular uniformity. This is especially important for implanting the bottom of high aspect ratio trenches and vias that are common in memory and logic devices.

2:30 PM L1-S3.2
SIMS Study of Boron Enhanced Hydrogen Diffusion in Amorphous Silicon Formed by Ion Implantation. (#891) Brett Cameron Johnson¹, Armand J. Atanacio², Kathryn E. Prince², Jeffrey C. McCallum¹; ¹ARC Centre of Excellence for Quantum Computer Technology, School of Physics, The University of Melbourne, Parkville, Victoria, Australia ; ²Australian Nuclear Science and Technology Organisation, Australia.

Hydrogen diffusion in amorphous silicon (a-Si) plays an important role in the behavior and types of defects present during device fabrication. The presence of H can also severely affect the efficiency of certain processing steps such as crystallization via solid phase epitaxy (SPE) where the SPE rate can decrease by up to 50%. In this instance, H infiltrates from the surface native oxide into the surface amorphous layer during thermal processing. Extensive H diffusion studies have been performed with hydrogenated amorphous silicon (a-Si:H) where H is known to reduce the dangling bond density allowing dopant activation suitable for photovoltaic devices. However, a-Si:H and a-Si formed by ion implantation are quite different materials. Indeed, the activation energy for H diffusion in intrinsic a-Si is about twice that found in undoped a-Si:H. Therefore, it might be expected that a different diffusion mechanism is operative in a-Si. There is some evidence that the H diffusion is governed by the dangling bond type defect. Coincidently, this mechanism may also be the basis of the SPE process mentioned above. The study of H diffusion in a-Si formed by ion-implantation is thus two-fold: To maximize the efficiency of processing steps of device fabrication and to determine whether links between H diffusion and SPE can be substantiated. In the present study, the evolution of a H implanted profile after partial anneals is followed using secondary ion mass spectrometry (SIMS). Multiple energy Si implantations into Si(100) wafers were used to form ~4.1 μm thick surface amorphous layers. The samples were then implanted with multiple energy B ions resulting in a constant concentration profile between 1.6 and 2.3 μm to peak concentrations of 1 x 10²⁰ and 1.5 x 10²⁰ B/cm³. A single 205 keV H implantation was then performed which had a projected range of 1.9 μm placing the H concentration profile in the middle of the B implanted region. Post implantation partial anneals were performed for various times in air over the temperature range 380 - 640 ^oC. The depth profiles of the B and resulting H concentrations were determined with a Cameca IMS-5F secondary ion mass spectrometer equipped with a Cs ion source. A primary ion beam of Cs+ with a net impact energy of 14.5 keV was applied for the depth profiling. The Cs⁺ beam was rastered over an area of 250 μm². Negative secondary ions were collected from the central part of the rastered region (55 μm in diameter) in order to avoid crater side wall effects. The SIMS detection limits were 2 x 10¹⁷ and 5 x 10¹⁵ for H and B, respectively. The enhanced diffusion coefficients are determined and it is found that the enhancement is greatest at the lowest temperature. At 460 ^oC the diffusion coefficient is enhanced by a factor of 10.1 and 16.5 for the 1 x 10²⁰ and 1.5 x 10²⁰ B/cm³ implanted samples, respectively. The similarities between dopant enhanced H diffusion and SPE are discussed and possible models for H diffusion are considered in light of these results.

2:45 PM L1-S3.3
Electronic Raman Spectroscopy of Ion Implanted Phosporus Donors in Silicon. (#1029) Paul Gregory Spizzirri, Jeffrey C McCallum, Steven Prawer; ARC Centre of Excellence for Quantum Computer Technology, School of Physics, The University of Melbourne, Parkville, Victoria, Australia.

Raman spectroscopy is well known for its ability to study lattice vibrations in semiconductors (i.e. phonons) however, it can also be used to probe the electronic states associated with donors, acceptors and conduction band electrons through bound electron transitions and free particle scattering. In the case of donors, isolated (i.e. non-interacting) states can be readily identified by their characteristic 1s(A) to 1s(E) absorption. When donor concentrations increase (i.e. donor states interact), the Raman transition broadens and shifts asymmetrically to lower energies. At the high concentration limit (i.e. [P] > MOTT metal-insulator transition), the material is characterized by impurity band conduction which manifests itself as a broad spectral continuum. It has been proposed that the observed changes arise from donor-donor interactions with energy splittings indicative of the exchange coupling constant J for [P] < MOTT. These transitions have also been shown to be very sensitive to the local donor environment (e.g strain fields, ionisation and surface charge). In the solid state realization of a silicon quantum computer using phosphorus (donor) qubits, low energy ion implantation is employed using the "Top-Down" fabrication strategy. Efficient dopant activation that is not dependent upon the incident ion energy (i.e. resultant ion depth) is a prerequisite. Issues such as dopant gettering at the silicon-oxide interface, reduced dopant activation and ionization by interfacial trap states could all impact upon the successful fabrication and operation of any such device. In this work, we report on the application of electronic Raman spectroscopy to the measurement of phosphorus ensembles that have been implanted (with subsequent RTA annealing) using various implantation strategies.

3:00 PM L1-S3.4
Puzzling Growth of Cavities During Annealing After High Energy and Fluence of Si-Ion-Implantation Carried Out for USJ Processing in Si. (#833) Mariaconcetta Canino¹, Gabrielle Regula¹, Maryse Lancin¹, Ming Xu¹, Bernard Pichaud¹, Esidor Ntzoenzok²; ¹Institut Matériaux Microélectronique et Nanosciences de Provence (IM2NP), CNRS, Marseille, France ; ²Conditions Extrêmes et Matériaux : Haute Température et Irradiation, CNRS, Orléans, France.

Two series of n-type doped Si samples A and B were implanted with Si ions at high dose (1x1016 cm-2) and at high energy (0.3 and 1.0 MeV, respectively) to i) preamorphize the sample surface; ii) improve the electrical activation of the boron implant iii) create a vacancy super saturation to stand the formation of bulk open volumes during the next two steps of the USJ manufacturing. They consist of implanting He either at 10 keV or 50 keV to create a self interstitial barrier diffusion (porous layer), and eventually, B atoms. Then, the samples undergo a rapid thermal annealing (RTA) at 900?C or 1000?C for 20s to grow He cavity layer close to the surface, embedded between the projected range (Rp) of boron and silicon. After the triple ion implantation, the associated junction depth Xj which depends on the occurrence of the so-called transient enhanced diffusion (TED) is measured by secondary ion mass spectroscopy (SIMS) before and after annealing. The value of Xj is related to the ability of the porous layer to trap self interstitial atoms. Cross section transmission electron microscopy (XTEM) studies on as-implanted samples A and B show the formation of an amorphous layer 570nm deep under the surface and between 700 nm and 1400 nm under the surface, respectively, as well as a damaged zone (but still crystalline ) at the Rp (projected range) of He (of course, only in the B samples). The XTEM observations are in accordance with the Rp of Si and He given by TRIM computing. In B samples, after 900?C RTA, in addition to the formation of a narrow band of small cavities or bubbles at Rp(He)=100nm or 400nm, a larger and deeper zone located at the Rp(Si), highly defective, poly-nanocrystalline and containing large cavities (about ten times bigger than the ones at Rp(He)) can be observed when Rp(He) is the deepest. Whereas the growth of cavities in the shallowest zone was intentional, the deepest one was actually unexpected and to our knowledge, never reported in the literature. Indeed, neither cavities at Rp(Si) or poly-nanocrystalline band are never observed in A samples, whatever the value of Rp(He). Consequently, though the poly-nanocrystalline structure of the previously amorphous region in B samples after annealing could be caused either by SPER (solid phase epitaxy regrowth) from both crystalline borders of the amorphous region, it is rather likely due to the presence of bulk open volumes which could play the role of germs for the nucleation and growth of Si nanocrystals. In this case, the re-crystallization mechanism should have an activation energy lower than the SPER one (2.7eV). Critical condition for the production of stable bubbles embryos is the presence of He. Therefore, though He is supposed to desorb beyond 500?C, a non negligible amount may actually diffuse towards the bulk (provided that it is not too close to the surface) to stabilize the vacancies created by Si implantation. After RTA at 1000?C the density of cavities both at the Rp(He) and at the Rp(Si) is significantly reduced in accordance with SIMS boron depth profiles, displaying a huge transient enhanced boron diffusion. The mechanism and kinetics of the open volume formation at Rp(Si) is still unclear: High dose Si implantation is going to be performed alone, in order to address the role of He in the formation of both cavities and poly-nanocrystals at Rp(Si). This work is funded by the French ANR "Nanocafon" program.

AFTERNOON BREAK 3:30 PM - 4:00 PM

4:00 PM *L1-S4.1 (invited)
Proton Beam Writing: Nanolithography Using a Focused Beam of MeV Protons. (#781) Andrew A. Bettiol, Jeroen A. van Kan, Ee Jin Teo, Mark B. H. Breese, Frank Watt; Department of Physics, National University of Singapore, Singapore.

Proton Beam Writing (PBW) is a direct write lithographic technique that utilizes a highly focused beam (~50 nm) of mega-electron volt (MeV) protons to pattern or modify a material or resist. PBW is well suited to producing high aspect ratio, sub-100 nm structures. This is due to the fact that the secondary electrons generated by the primary proton beam are of the order of a few kilo-electron volts or less in energy. Therefore there is very little delocalization of the energy deposited by these secondary electrons, and proximity effects are minimal. Furthermore, the relatively large penetration depth of MeV protons allows for high aspect ratio structures to be fabricated. PBW can be used to pattern or modify a wide variety of materials including polymers, insulators and semiconductors. Depending on the application, the proton beam modified material may undergo additional processing steps such as etching or electroplating in order to form the final structure. In this paper we will compare the PBW technique with existing direct write lithographic techniques, outlining the unique features and advantages of using MeV protons for nanolithography. In addition, various applications of PBW will be presented including applications in microphotonics, micro-fluidics, lab-on-a-chip devices and metamaterials.

4:30 PM L1-S4.2
Ion Beam Synthesis: Predicting Nanocluster Size Distributions. (#861) Daryl Chrzan¹, C. W. Yuan¹, S. Shin¹, Julian Guzman¹, C. Y. Liao¹, Ian D. Sharp², Joel W. Ager³, E. E. Haller¹; ¹Materials Science and Engineering, University of California, Berkeley and Lawrence Berkeley National Laboratory, USA ; ²Walter Schottky Institut, USA ; ³Lawrence Berkeley National Laboratory, USA.

Ion beam synthesis is a promising production route for generating a wide variety of embedded nanoclusters due to its chemical flexibility and its compatibility with present day semiconductor processing routes. However, the technique often produces cluster size distributions that are broad - the width of the size distribution is typically comparable to the average cluster size. It is desirable, therefore, to develop theories capable of predicting quantitatively the size distributions of clusters fabricated using IBS, in order that such theories might be used to identify key parameters determining cluster size distributions. We have developed two distinct theoretical approaches to predict the size distributions arising during IBS: kinetic Monte Carlo simulations, and the mean-field self-consistent solution to a set of coupled rate equations. Remarkably, these markedly different approaches produce size distributions with shapes that not only agree quantitatively with one another, but also agree quantitatively with experimental observations. Further, the theories allow identification of the parameters that influence the shape of the size distribution, and lend insight into how one might alter the observed size distributions. Both approaches suggest that the growth of clusters during implantation proceeds in stages. In the first stage, clusters nucleate and begin to grow, much as islands nucleate on a surface during molecular beam epitaxy. The average cluster size increases during this stage of growth, while the cluster number density remains low. Eventually, however, the clusters reach sizes at which they are very likely to be hit directly by a deposited ion. These collisions lead to fragmentation of the clusters. Thus, this second stage of growth is marked by a decreasing average cluster size, and an increasing cluster number density. During the last growth stage, the shape cluster-size distribution can become invariant. The invariant shape of the cluster distribution is governed by a small set of parameters. The most important of these is the ratio, F/D*, of the ion flux, F, to the effective ion diffusion coefficient of the implanted species, D*. Predicted distribution shapes agree quantitatively with experimentally observed distributions for Ge, Co and Ag embedded in silica. This research is supported by the Director, Office of Science, Office of Basic Energy Sciences, Division of Materials Science and Engineering of the U. S. Department of Energy under contract DE-AC02-05CH11231

4:45 PM L1-S4.3
Ion-Beam Synthesis of Silica Nanostructures. (#1336) Avi Shalav, Robert G. Elliman, Taehyun Kim; Electronic Materials Engineering Department, Research School of Physical Sciences and Engineering, The Australian National University, Australian Capital Territory, Australia.

A random array of silica nanowires can be grown on silicon by metal-induced vapour-liquid-solid (VLS) growth. Surprisingly this can be achieved by simply depositing a thin metal film on silicon and annealing it in Ar or N₂ at 1100^oC. In this case the metal 'catalyst' particles are formed by decomposition of the thin film into small islands. The wires grow from these particles to form a dense, random arrangement of the wafer surface. However, the mechanism by which this occurs has been the subject of considerable speculation and debate. The resulting nanowires are amorphous SiO_x (with x less than 2) and typically have diameters in the range 10-1000 nm and lengths in excess of 100 mm. Their large surface-volume fraction together with their inertness and biocompatibility makes them attractive functional materials for applications in photonics and environmental and biosensing.

In this study we show that ion-implantation is an effective alternative to deposited thin films for promoting the nucleation and growth of nanowires and that it affords additional flexibility for fabricating functional nanowire structures. As a means of forming the initial catalyst ion implantation is shown to have the advantage of premixing the catalyst metal with silicon to assist in the formation of the active eutectic or silicide phase, and to leave the silicon surface exposed to the annealing ambient which promotes nanowire growth. The significance of these features is discussed in terms of the nanowire growth mechanism. The dependence of the nanowire morphology (diameter and length) on deposition and implant conditions is also discussed.

5:00 PM L1-S4.4
Waveguide Fabrication in Bulk Silicon Using Focused Proton Beam Irradiation. (#890) Ee Jin Teo¹, Andrew Anthony Bettiol¹, Mark Breese¹, Pengyuan Yang², Goran Mashanovich², Bill Headley², Graham Reed²; ¹Centre for Ion Beam Applications, National University of Singapore, Singapore ; ²University of Surrey, United Kingdom.

In this paper, we present our recent work on proton beam writing of waveguides in bulk silicon. A high energy focused beam of 250 keV protons is used to selectively scan across a p-type silicon wafer. This increases the local resistivity of silicon, and reduces the rate of porous silicon formation in the irradiated region during subsequent electrochemical etching in hydrofluoric acid. By undercutting the irradiated structure, it is possible to produce a silicon core that is surrounded by porous silicon cladding. We show that the etch rate is strongly dependent on the irradiated dose, enabling us to have three-dimensional control of waveguide core. Using this direct-write process, it is possible to build up any pattern of accumulated damage to create the desired waveguide profile in a single irradiated step. This is potentially important for fabrication of silicon waveguide taper. Propagation loss of the proton beam written waveguide is measured at 1550 nm using cut-back method. It is found that annealing of the waveguide at 500?C is important to remove the proton-induced defects and reduce the waveguide loss. As the porous silicon cladding layer can be easily removed with diluted potassium hydroxide, this technique allows the flexibility to fabricate various types of waveguides, such as ridge, channel and air-cladded waveguides.

L1-S5.1
Si PIN detectors for keV Ions: Characterisation of Detector Damage Using MeV IBIC. (#992) Andrew David Charles Alves, Changyi Yang, David Jamieson; ARC Centre of Excellence for Quantum Computer Technology, School of Physics, The University of Melbourne, Parkville, Victoria, Australia.

To detect a single low energy (a few keV) P ion is a critical aspect of the top down approach to the fabrication of a Si quantum computer. A Si PIN detector operating with extremely low noise and 5nm oxide has been developed able to produce an electronic pulse from a 14 keV P ion impact [1]. The charge collection efficiency of these detectors is routinely compared against commercial PIN detectors using 2 MeV He ion beam induced change IBIC measurement. This MeV IBIC characterisation is an important part of the quality control in detector fabrication. This study extends the MeV IBIC characterisation to include measurement of the damage rate as a function of ion fluence in different detector regions. This data is extremely important for applications that may require the implant of large arrays of individually detected P ions because detector damage may become a critical impediment to detector performance. The results showed a maximum of 2 % decrease in charge collection efficiency when the detector was exposed to 10⁶ He ions in a 30x30 μm² area. [1] C. Yang et. al., Jpn. J. Appl. Phys. (2003) 4124.

L1-S5.2
Ion Beam Lithography-Optimization for Patterning Including Few-Nanometer Features. (#1213) John E.E. Baglin, A. J. Kellock; IBM Almaden Research Center, San Jose, California, USA.

Patterned ion beam exposure of an appropriately tailored resist should in principle be capable of supporting full-chip lithography of arbitrary multi-scale patterns that include feature sizes down to a few nanometers, with minimal line edge roughness and negligible proximity effects, in a process with high reliability, high dimensional stability, minimal defects, and high throughput. Due to many factors unique to ion beams, such as the absence of interference issues, the controlled depth of ion penetration, the option to inhibit charge buildup on the patterned device, and the variety of beam parameters available to best suit each case, ion beam lithography offers simple alternatives to existing approaches, appropriate for device fabrication beyond currently planned generations of CMOS processing. Even if we assume a perfectly patterned beam, such as might be produced by ion beam projection systems, a key challenge remains - selecting optimal parameters for capturing such a pattern with precision in an optimally designed lithographic medium. In this presentation, we will analyze the issues that need to be addressed, and discuss experimental attempts to resolve them by suitable choice of resist material, molecuar weight, and thickness; ion species, energy and fluence; substrate temperature; and surface smoothing. Our experimental examples mostly involve broad-beam exposure using stencil masks, of various resist materials on silicon or on deposited metal surfaces. Ion species include H+, He+, Ne+, Ar+, at energies ranging from a few hundred eV to 300 keV, exposing a variety of resist coatings offering high spatial resolution. SRIM based simulations of ionization cascades in the resist layer are used to support an exposure model that addresses the severe limitations of shot noise and line edge roughness, and uniformity of exposure through the depth of the resist layer. We conclude that, with the use of suitably customized resists, and optimal ion beam parameters, few-nm feature resolution can realistically be expected.

L1-S5.3
Strategies for Fabricating Nanostructures with Focused Ion Beam Technology. (#701) Vishwas Jagannath Gadgil, MESA+ institute for Nanotechnology, University of Twente, Enschede, Netherlands.

Abstract Focused ion beam has emerged as an important tool in research in the field of nanotechnology. Specialized equipment is now commercially available with a built in scanning electron microscope for observing the fabricated structures (1). With computer controlled ion beam movement, it is possible to mill complex patterns such as photonic devices (2). It is possible to control parameters such as beam currant, spot overlap, dwell time and milling sequence. Patterns can be programmed and run on the systems as stream files. One of the problems that limits the accuracy and resolution of the patterns being milled is the re deposition. Depending on the milling parameters, re deposition can take place in varying degrees. In this investigation, the effect of the milling parameters on the fabrication of patterns in silicon is evaluated. Another problem with milling patterns is that the beam is not switched off while traveling from one feature in the pattern to the next. This means that with high number of loops the surface between the features is also milled. This makes it necessary to find an optimum between the dwell time which influences re deposition on one hand and the number of loops which can lead to unwanted milling on the other (3). However with new developments, it is now possible to control the beam blanker with the stream files, which eliminates the unwanted milling. The effect of beam blanker on milling of patterns is investigated. Results of the milled patterns are presented. Cross sections are investigated to evaluate extent of re deposition. Milling strategies for optimum results are discussed. 1. Applications of Focused Ion Beam in Nanotechnology, V.J. Gadgil, F. Morrissey, Encyclopedia of Nanoscience and Nanotechnology, H.S. Nalwa ed. Vol. 1: Pages 101-110, American Science Publishers, 2004 2. Fast prototyping of planar photonic crystal components using a combination of optical lithography and focused ion beam etching', Cazimir G. Bostan, Ren? M. de Ridder, Vishwas J. Gadgil, Henry Kelderman, Alfred Driessen1, Laurens Kuipers, 12th European Conference on Integrated Optics ECIO, Grenoble France April 6-8, 2005, (proceedings). 3. Focused Ion Beam Milling Strategies of Photonic Crystal Structures in Silicon, Wico C.L. Hopman, Feridun Ay, Wenbin Hu, Vishwas J. Gadgil, Laurens Kuipers, Markus Pollnau, Ren? M. de Ridder, European Conference on Integrated Optics, April 25-27, 2007, Copenhagen, Denmark (conference proceedings)

L1-S5.4
Nanopillar Multiple Quantum Wells Fabrication by Focused Ion Beam and Application to Atom Probe Tomography Specimen Preparation. (#611) Baptiste Gault, Australian Key Centre for Microscopy and Microanalysis, Australia.

B. Gault1*, S.-E. Wu2, C.-P. Liu2, T.-H. Hsueh, S.P. Ringer 1 1-Australian Key Centre for Microscopy and Microanalysis, The University of Sydney, NSW 2006, Australia 2-Department of Materials Science and Engineering, National Cheng-Kung University, Tainan 70101, Taiwan The advent of laser assisted atom probe tomography (APT) [1, 2] has opened the technique to the analysis of poorly conducting materials. The main issue has been to find ways to prepare specimens, which must be in the form of a very sharp needle, with a radius of curvature at the apex smaller than 100 nm. This is an important problem of relevance to electronic materials because the atom probe offers the possibility to map the distribution of individual atoms in 3D, with a sub-nanometre resolution. Therefore, APT is emerging as one of the most promising techniques available for the study of structure-function relationships in these materials. Moreover, the coupling between a focused ion beam (FIB) and in-situ micromanipulators has enabled the fabrication of site-specific specimens [4]. Very recently, Wu et al. [5] have developed a self-masking method to prepare nanoscale pillars containing InGaN/GaN Multiple Quantum Wells in GaN wafers using a FIB based technique. In this presentation, we will discuss the fabrication of these nanopillars in the FIB, and their potential use as specimen for APT and the implications of this for nanometrology in electronic materials. [1] G.L. Kellog and T.T. Tsong, J. App Phys 51, 1184 (1980). [2] B. Gault et al., Rev Sci Instrum 77, 043705 (2006). [2] D.J. Larson et al., Ultramicroscopy 79, 287 (1999). [3] M.K. Miller et al., Ultramicroscopy 102, 287 (2005). [4] T. F. Kelly et al., Ann. Rev. Mat. Res. 37, 681-727 (2007). [5] S.-E. Wu et al., Nanotechnology 18 445301 (2007) .

L1-S5.5
Production of Pores in Fluoropolymer Membranes by Ion Beam Bombardment. (#1210) Renato Amaral Minamisawa, Robert Lee Zimmerman, Daryush Ila; Center for Irradiation of Materials (CIM), Department of Physics, Alabama A&M University, USA.

Producing structures in membranes at the nanometer scale can serve several applications such as to localize molecular electrical junctions and switches, and to function as masks. In previous work we demonstrated the fabrication of porous membranes in masked fluoropolymer films using scanned ion beam bombardment. The process dispenses the use of time consuming chemical and etching processes. Here we report on the creation mechanism of pores using ion bombardment. Aspects of the ion beam interaction with matter are explained as well as an analysis of the shape of the fabricated structures. The pores were produced using our feedback controlled ion beam apparatus and were analyzed using optical and atomic force microscopic (AFM) analyses.

L1-S5.6
Control of Microbial and Cell Growth on Biomaterials by Silver Ion Implantation. (#1211) Daryush Ila, Robert Lee Zimmerman; Center for Irradiation of Materials (CIM), Department of Physics, Alabama A&M University, USA.

High energy ion beam bombardment and Ion Beam Assisted Deposition (IBAD) of trace amounts of elements near the surface of biocompatible metals, polymers and ceramics significantly modify their mechanical, electrical and chemical properties. Medical implants usually require seamless adaptation with adjacent living tissue. For example, temporary carbon trans coetaneous electrodes or drainage tubes interface with skin. Permanent replacements of artificial teeth and the femur head interface with bone of the jaw and femur, respectively. We have shown that ion beam induced roughening of specific surfaces of these implanted materials enhance the adhesion with adjacent living tissue. We have shown that implantation of silver ions below the surface of otherwise biocompatible materials completely inhibits cell attachment, while leaving neighboring areas hospitable to cell attachment and adherence. This patterning permits precise control of the formation of tissue on implanted biomaterials. For example, the moving parts of carbon heart valves are sites for cell attachment, potentially dangerous if the resulting tissue later detaches. Surface treatment by silver ion implantation eliminates tissue formation. Similarly, we have shown that silver implantation at the surface of Ultra High Molecular Weight Polyethylene (UHWMPE) gives the material anti microbial properties, as well as increasing the wear resistance. Both are improvements to the UHWMPE liner for the acetabular cup replacement for the hip joint.

L1-S5.7
Raman Scattering Study of Defects in H-Implanted Si. (#889) Brett Cameron Johnson, Jeffrey C. McCallum; ARC Centre of Excellence for Quantum Computer Technology, School of Physics, The University of Melbourne, Parkville, Victoria, Australia.

The evolution of H-related defects during implantation and annealing has been extensively studied using Raman scattering spectroscopy. A number of lines between 1800 and 2400 cm^-1 have been attributed to Si-H bond vibrational frequencies. The appearance and disappearance of these lines are dependent on the anneal temperature. Single 30 keV H ions were implanted into Czochralski Si (100) wafers held at temperatures of 195, 25 or 350^oC. The samples were then cut up into pieces before undergoing post-implantation anneals between 150^oC and 600^oC for 30 minutes in an Ar ambient. Each sample was annealed only once. The interaction of H-related defects with pre-existing nanocavities was also studied during H-implantation and anneal processing steps. The evolution of H-related defects was followed using Raman spectrometry. Spectra were collected with a single grating Renishaw RM1000 Raman system. The 514.5 nm line from a continuous wave Ar⁺ ion laser was scattered from the sample in a backscattering geometry and spectrally resolved onto a CCD camera. In the present paper, we show that the H-related defects produced during implantation have a strong dependence on the implantation temperature. A number of features are absent in the Raman spectrum of samples implanted at high temperatures suggesting that the evolution of H-related defects in c-Si is heavily dependent on the dynamics of defect diffusion during implantation. The evolution of defects during post-implantation anneals also has a dependence on the initial implant temperature The interaction between H-related defects and pre-existing nanocavities is apparent in the Raman spectra. The nanocavities show strong defect gettering behavior as the implant temperature is increased.

L1-S5.8
Intrinsic and Dopant-Enhanced Solid Phase Epitaxy in Amorphous Germanium. (#963) Brett Cameron Johnson, Jeffrey C. McCallum; ARC Centre of Excellence for Quantum Computer Technology, School of Physics, The University of Melbourne, Parkville, Victoria, Australia.

Due to recent developments in nano-scale electronics and opto-electronic devices, Ge has regained some interest. It also provides an alternative system to Si to test various processes and models. Solid phase epitaxy (SPE) is a central process in device fabrication and is used to activate dopants implanted into an amorphous layer. SPE provides a means of achieving high dopant activation with a low thermal budget. At present there are only a limited number of studies on SPE in amorphous Ge (a-Ge). This work presents a comprehensive set of measurements of intrinsic and dopant-enhanced SPE for a-Ge layers formed by ion implantation on <100> Ge substrates. A number of Ge substrates from different suppliers and having various background doping concentrations were used. Samples were annealed in air at temperatures in the range 300-540 ^oC and the rate of interface motion was monitored using time-resolved reflectivity. SPE rates showed an Arrhenius behavior with an activation energy of (2.15 ± 0.04) eV and velocity pre-factor of (2.6 ± 0.5) x 10⁷ m/s. SPE rate measurements on both thick (3.25 μm) and thin (1.5 μm) a-Ge layers have been performed to distinguish between bulk and near-surface SPE growth rate behavior. Explosive crystallization was a common occurrence in the thick amorphous layers and limited the SPE anneal to temperatures below 440 ^oC. The source point for the explosive crystallization event was often a small chip in the cleaved edge of the sample. Hydrogen infiltration from the surface into the amorphous layer is observed by secondary ion mass spectroscopy. The infiltration is not as dramatic as that found in a-Si possibly due to the instability of the oxide on the Ge surface. Hydrogen is shown to affect the SPE rate up to 0.3 μm into the layer independent of the initial amorphous layer thickness. This may have affected all previous SPE measurements in a-Ge. Dopant enhanced SPE rates were also measured in a-Ge layers containing uniform concentration profiles of As or Al formed by multiple energy implantation. Concentrations spanned the regime 1-10 x 10¹⁹ /cm³ and were found to result in enhancements similar to those found in Si based SPE experiments. Dopant compensation effects were also observed in a-Ge layers containing equal concentrations of As and Al, where the SPE rate is similar to the intrinsic rate. The validity of various SPE models is tested with our data. The extended kinetic model provides good fits to the data and a refined value of the number of crystallization events per formation of a dangling bond pair was found. Likewise, the generalized Fermi level shifting model showed excellent fits and compared well to fits to similar data obtained for Si.

L1-S5.9
Ion Implantation-Based Patterning for Nanoparticle Assembly. (#1120) Naoki Kishimoto¹, Jin Pan², Masahide Nakamura², Yoshihiko Takeda¹; ¹Quantum Beam Center, National Institute for Materials Science, Sakura, Tsukuba, Ibaraki, Japan ; ²University of Tsukuba, Tsukuba, Ibaraki, Japan.

Ion beam-based techniques offer various possibilities for robust spatial control of nanostructures, either in self-assembled- or in actively controlled manners. Possible approaches are not only atom-supply control (e.g., masked implantation, IPL) but also perturbation control of fields interactive with implants (e.g., photon, phonon, mechanical fields). We have explored lateral control methods for nanoparticle assembly by perturbing photon- and stress/strain fields, as well as masked implantation. Patterning of metal nanoparticles embedded in insulators is one of the most important targets for plasmonic applications. However, the nanoparticle patterning is subjected to inherent problems, i.e., necessity of high fluences, inaccuracy due to ion straggling and indispensable rearrangement of implants. Here, we present a few methods of nanoparticle patterning. Ion implantation into amorphous SiO₂ was conducted with 60 keV Cu^-. The perturbations of interest are laser and nanoindentation as photon and stress/strain fields, respectively. Simultaneous laser irradiation under ion implantation either enhanced or decreased surface plasmon resonance, i.e., nanoparticle precipitation/growth, depending on ion fluence, photon energy and laser power. Nano/micro-indentation in periodic arrays enhanced nanoparticle precipitation under Cu ion implantation. The defect-implant interaction leaded to control of nanoparticle precipitation. Masked implantation with either a lithographic mask or a porous-alumina membrane was also conducted. The perturbations interactive with nanoparticle precipitation/dissolution can be used for controlling nanoparticle assembly. The problems for patterning nanoparticle assembly will be discussed.

L1-S5.10
Hydrogen Diffusion and Trapping in Ge and GeSi Alloys for Wafer Cleaving. (#1217) Daniel J Pyke¹, Jeffrey C McCallum², Robert G. Elliman¹; ¹Electronic Materials Engineering Department, Research School of Physical Sciences and Engineering, The Australian National University, Australian Capital Territory, Australia ; ²School of Physics, The University of Melbourne, Parkville, Victoria, Australia.

With the advent of the ion-cut procedure, wherein hydrogen ions are implanted into a silicon wafer and annealed to induce microcracks that cleave the wafer, there has been intense interest in understanding hydrogen diffusion and trapping in silicon. Significantly, recent studies have shown that hydrogen accumulates in regions of high compressive stress caused by the implant damage rather than at the peak of the initial implant distribution, and this has led to the possibility of using a strained layer, such as an epitaxially-grown GeSi layer, to control the cleavage process. Extension of the ion-cut process to other semiconductors is also highly desirable. With this in mind, the present study examined hydrogen induced defects, diffusion and trapping in Ge and GeSi alloys. The latter included samples with well defined strain profiles produced by varying the Ge composition in an epitaxial GeSi alloy layer grown on (100) silicon. The microstructure of hydrogen implanted samples was characterised by Raman spectroscopy and both plan-view and cross-sectional transmission electron microscopy (TEM), whilst strain distributions were determined by double crystal x-ray diffraction. Composition analysis of alloy layers was performed by Rutherford backscattering spectrometry, and hydrogen profiles were determined by elastic-recoil detection (ERD) and secondary ion mass spectrometry (SIMS). Hydrogen induced defects in Ge and the effect of strain and strain gradients on hydrogen diffusion and trapping is reported.

L1-S5.11
Calibration of Secondary Ion Mass Spectrometry of RF - Sputtered Indium Nitride Thin Films with Elastic Recoil Detection Analysis. (#824) Saravanan Somasundaram, Heiko Timmers, Santosh K Shrestha, K. Scott A. Butcher, Marie Wintrebert-Fouquet, Armand J Atanacio; Cell Technology, TATA BP Solar India Ltd, Bangalore, India.

S. Saravanan1, Heiko Timmers1, Santosh K. Shrestha1, K. Scott A. Butcher2, Marie Wintrebert-Fouquet2, Armand J. Atanacio3 1School of Physical, Environmental and Mathematical Sciences, University of New South Wales at the Australian Defence Force Academy, Canberra, ACT 2600, Australia 2Department of Physics, Macquarie University, Sydney, NSW 2109, Australia 3Australian Nuclear Science & Technology Organization, Sydney, NSW 2234, Australia Indium Nitride is an important III-V semiconductor with many potential applications in optoelectronic devices. The growth of Indium Nitride thin films has started in the 1970's. In the 1980's extensive work on Indium Nitride thin films has been carried out due to their specific properties like smaller band gap energy, carrier transport characteristics etc. In this work, Indium Nitride thin films of two different thicknesses were fabricated on glass substrate by RF reacting sputtering of an Indium target with pure nitrogen gas. These Films were characterised by using x ray diffraction (XRD) and Secondary Ion Mass Spectrometry (SIMS). XRD pattern shows that the InN films grown are well oriented with all crystallites having (0002) and (0004) planes parallel to the growth surface. SIMS reveals that both Indium and Nitrogen are distributed uniformly with in the epilayer. The SIMS data is calibrated with Elastic Recoil Detection Analysis (ERDA) and the results are correlated. Key words: Indium Nitride, SIMS, Ion Beam Analysis, Thin films

L1-S5.12
Advances in the Precision Implantation of Single Ions. (#968) Andrew Alves¹, Jessica Van Donkelaar¹, Changyi Yang¹, Alberto Cimmino¹, Sergey Rubanov², David Jamieson¹; ¹ARC Centre of Excellence for Quantum Computer Technology, School of Physics, The University of Melbourne, Parkville, Victoria, Australia ; ²Bio21 Institute, The University of Melbourne, Parkville, Victoria, Australia.

The top down approach to the fabrication of a single atom device requires the detection and precise positioning of each ion in the substrate. In Si we have seen advances in detection using charge collection [1] and positioning using the nanostencil technique [2]. This study reports on the advances made towards developing a nanostencil apparatus at the ARC Centre of Excellence for Quantum Computer Technology, University of Melbourne node. Initial proof-of-principle experiments have been performed and the technical details are explained here. Slotted nanostencils in Si cantilevers have been fabricated using a Ga focused ion beam with widths down to 20 nm. Ion beam induced charge mapping with a beam of 1.5 MeV He ions confined by a 500 nm wide nanostencil have been performed. Nanoscale lithography in PMMA resist with both 1.5 MeV He and 14 keV Ar ions is used to reveal the ion confinement that can be achieved with the nanostencil. [1] C. Yang et. al., Jpn. J. Appl. Phys. (2003) 4124. [2] H. Guo et. al., Applied Physics Letters 90 (1999) 093113.

L1-S5.13
Nucleation and Growth of Gold Particles on the Surface of Gold-Implanted Silicon. (#1402) D.K. Venkatachalam¹, D.J. Llewellyn², K.B. Belay², Robert G. Elliman², D.K. Sood¹, S.K. Bhargava¹; ¹RMIT University, Melbourne, Victoria, Australia ; ²Electronic Materials Engineering Department, Research School of Physical Sciences and Engineering, The Australian National University, Australian Capital Territory, Australia.

The use of metal nanoparticles as catalysts for the growth of nanowires is well established and such particles can readily be formed by the decomposition of deposited thin films or chemically in colloidal form. It has recently been shown that such particles can also be formed by ion-implanting silicon with high-fluences of metal ions and annealing to produce precipitates on the silicon surface. This process has the advantage of spatial selectivity as well as the ability to combine elements and/or employ single isotopes in the catalyst particle. In preliminary studies we showed that ordered circular patterns of gold nanoclusters can also be produced under specific implantation and annealing conditions. In this work, we have extended these studies by performing in-situ annealing experiments in both scanning- and transmission -electron microscopes, as well as by performing ex-situ measurements of samples subjected to short-time anneals using rapid-thermal annealing (RTA). For these experiments Si (100) single crystal substrates were implanted at room temperature with 10 keV Au²⁺ ions to fluences up to 4x10¹⁶ Au/cm^-2 followed by rapid thermal annealing at temperatures in the range from 450°C - 750°C and for durations of 5-30s. The evolution of the microstructure is shown to be consistent with a model in which the implanted gold forms a liquid Au-Si eutectic layer before segregating to the surface and precipitating as nanoparticles. These observations, together with data on the formation of ordered particle arrays are presented.

L1-S5.14
IBIC Characterization of Micro-JFET Device for Single Ion Implantation Application. (#472) Changyi Yang, School of Physics, The University of Melbourne, Parkville, Victoria, Australia.

This research work explored the potential of accurately doping micro/nano-electronic devices with a novel single ion implantation fabrication method. Electronic devices such as JFETs with pre-fabricated drain and source structures were tested with Ion Beam Induced Charge (IBIC) analysis. These existing drain and source of the standard electronic device were used as detector electrodes and near 100% charge collection efficiency was achieved at the central area between the drain and source. The devices with high charge collection efficiency can be configured as single ion detectors for accurately placing designed numbers of single dopants including phosphorus and boron impurity ions under a control gate, aiming for the enhancement of the device operation reliability through suppressing the fluctuation of the dopant atom numbers in the conventional semiconductor doping process. A further development toward a nanometer scaled electronic device is outlined.

11:00 AM *L2-S2.1 (invited)
Instantaneous Analysis of Real-Time RBS Data Using Artificial Neural Networks. (#749) Dries Smeets¹, Nuno P. Barradas², Armando Viera³, Jelle Demeulemeester¹, Craig M. Comrie⁴, Chris C. Theron⁵, André Vantomme¹; ¹Instituut voor Kern- en Stralingsfysica and Institute for Nanoscale Physics and Chemistry, K.U. Leuven, Belgium ; ²ITN, Portugal ; ³ISEP, Portugal ; ⁴UCT, South Africa ; ⁵iThemba, South Africa.

Rutherford backscattering spectrometry (RBS) has proven most valuable in studies where the compositional depth profile of thin films is investigated as a function of thermal treatment. Often the depth sensitive information from RBS is complemented by the phase sensitive information of X-ray diffraction (XRD) for a complete overview of the response of a thin film to thermal annealing. Conventionally, several specimens are subjected to different heat treatments and subsequently analyzed one by one. A large number of samples and annealing treatments is required to determine, for example, diffusion coefficients or activation energies. It is much more convenient to determine the specimen properties in real time, i.e. during annealing. This does not only drastically decrease the workload: kinetic parameters can, for example, be obtained from a single ramped annealing; it also virtually eliminates the influence of small differences in annealing procedures and specimen preparation. These and other advantages of real time measurements have already led to the routine use of real-time XRD in thin film studies. Real-time RBS, on the other hand, suffers from the drawback that the analysis of the hundreds of spectra obtained in a typical experiment is very time consuming. It has therefore not yet found wide application. Artificial Neural Networks (ANNs) offer a solution to this shortcoming of real-time RBS. Using these networks, RBS spectra are analyzed instantaneously. Moreover, after building and training the ANNs, the analysis is performed without human interference. Although neural networks are capable of analyzing similar RBS spectra in a very short time, a new network must be built for each different system studied. This marks the disadvantage of ANNs for the analysis of RBS data that has so far prevented their widespread use. The combination of artificial neural networks and real-time RBS however, results in a perfect synergy and may announce the breakthrough of real-time RBS as well as ANNs for the analysis of RBS data. The huge number of spectra involved in real-time RBS experiments can be analyzed instantaneously with ANNs and the construction and training of neural networks pays off because of the large number of similar spectra that have to be analyzed. Since the same sample is analyzed, the consecutive spectra in a real time experiment do not differ that much, which is also true for different heat treatments. Moreover, more general networks can be trained to cover a set of real-time data (e.g. thickness or composition ranges). We will present the progress in real-time RBS credited to the routine use of ANNs. We will comment on the construction and training of neural networks for specific studies, e.g. studies of silicide growth kinetics, marker experiments and elemental redistributions during thin film reactions, and the advantages of real-time RBS over conventional annealing experiments.

11:30 AM L2-S2.2
Developments in Materials Characteration Using Ultra-Low Energy SIMS and Focused Ion Beam SIMS. (#1073) David Stuart McPhail, Richard Chater; Materials Department, Imperial College, London, United Kingdom.

The latest generation of SIMS depth profiling tools invoke ultra-low energy ion beam columns capable of producing reasonable primary beam currents at beam energies approaching 100eV. The ultra-low energy beams produce sub-nanometre depth resolutions and almost entirely eliminate the pre-equilibrium period in a SIMS depth profile. However there are combinations of beam energy and angle where ripple formation occurs almost instantaneously and these conditions may not be used. Thus the SIMS community is studying the scope of ultra-low energy SIMS depth profiling and the physics of ion beam-solid interactions that take place in this regime. In this paper I will discuss these trends and will give some examples of the applications of ultra-low energy SIMS with particular reference to dopant profiling in semiconductors. In another development a few Focused Ion Beam (FIB) tools have had secondary ion analysers added to their configuration leading to a new analysis option, FIB SIMS. In principle FIB SIMS offers SIMS with 5nm lateral resolution although the analytical volume is then very small, limiting sensitivity. We have applied our FIB SIMS tool to a number of systems and we have been conducting studies to see how the secondary ion yield may be improved. These studies of ion yield involve pre-dosing the surface with low energy ions, to change the surface chemistry. Applications of Secondary Ion Mass Spectrometry in Materials Science, D.S. McPhail, Invited review in the Journal of Materials Science, 40th Anniversary Issue, 41, pp. 873-903, Feb 2006. Applications of Focused Ion Beam SIMS in Materials Science, R.J. Chater and D.S. McPhail, Publication electronically in Mikrochimica Acta, http://www.springerlink.com/content/h535065176348842/, 22nd January 2008.

11:45 AM L2-S2.3
X-Ray Photoelectron Analysis of Ion Implanted and Ion Beam Mixed Polyetheretherketone. (#1208) Eric Loyd Tavenner¹, Barry Wood², Matt Curry³, Ryan Geidd³; ¹Ian Wark Research Institute, The University of South Australia, Australia ; ²Brisbane Surface Analysis Facility, The University of Queensland, Brisbane, Australia ; ³Center for Applied Science and Engineering, USA.

Conductive polymers were discovered by accident by Heeger et al in 1974 [1]. Since that time, electrically conductive polymers have been developed with the goal of replacing their inorganic counterparts in electronic devices, with the majority of work focused on developing these polymers by chemical processing. Ion implantation is another means of producing electrically conductive polymers. Unlike chemical processing, ion implantation can modify polymers in ways that are not achievable by chemical means, and is one of the few techniques that can change the characteristics of a material by non-equilibrium conditions. In 1982 the ion beam modification of polymers was first reported by Forrest et al [2], however this technique of producing electrically conductive polymers has not received the same level of attention as chemically produced polymers. This could be due to the complexities of characterising and elucidating the processes that occur in implanted materials. As a consequence of this, the ability to tailor the ion implantation procedure to a specific polymer to achieve desired characteristics is, to date, only attainable through trial and error. In addition, the study of ion implanted polymers has covered a wide range of polymer types with varied implantation parameters, with few common elements in the studies. This work looks at the XPS analysis of polyetheretherketone implanted by three different, but complimentary, techniques: nitrogen ion implantation, tin ion implantation and finally nitrogen ion beam mixing of thin tin films into the polymer. Three sample groups were made from 0.1 mm thick Polyetheretherketone films. One sample group was implanted with 50 kV nitrogen ions to a dose of 10¹⁶ ions/cm², another was implanted with either 10 kV or 45 kV tin ions to a dose of 10¹⁶ and 10¹⁷ ions/cm², and the last group was implanted in a similar manner as the first group but with a 100 Å thin tin or tin/antimony film thermally deposited before ion implantation. It was found that graphitic-like structures were produced within the implanted region. This is evident through the XPS C 1s peak low energy broadening and the formation of a high energy tail, and through the changing of the valence band to a graphitic line shape. Because of this, the C 1s line shape of graphite was used in the analysis of the implanted PEEK spectra, and it was discovered that all implanted samples had developed these structures. It was also seen through XPS that carbon-tin bonds were being formed during tin ion implantation and ion beam mixing of tin films, and that a significant amount of tin oxides were also present. Also of interest was the discovery that the conductivity of the tin mixed system can be made to go from insulating to metallic with the addition of a minimum amount of antimony. [1] J. Jagur-Grodzinski, Polymers for Advanced Technology 13, 615 (2002). [2] S. Forrest, M. Kaplan, P. Schmmidt, T. Venkatesan and A. Lovinger, Appl. Phys. Lett. 41(8), 708 (1982).

12:00 PM *L2-S2.4 (invited)
Medium Energy Ion Scattering for the Characterisation of the Ultra Shallow Implants and High K-Dielectric Gate Oxide Films. (#346) Jaap Albert Van den Berg¹, M A Reading¹, D G Armour¹, P Bailey¹, T.C Q Noakes²; ¹Joule Physics Laboratory, Institute for Materials Research, University of Salford, United Kingdom ; ²Science and Technology Facilities Council (CCLRC), Daresbury Laboratory, United Kingdom.

The continuing size reduction of transistors is transforming microelectronics into nanoelectronics as critical device dimensions are reduced to ~20 nm and below. The characterisation of shallow source/drain extension implants (e.g. dopant and damage profiles, layer regrowth and redistribution effects upon annealing) and especially of novel alloyed high-k gate oxide stacks (constituent profiles, ultra thin interfacial oxides, nitridation effects) represents a major analytical challenge. Medium energy ion backscattering (MEIS) is capable of yielding depth dependent information on the structure (e.g. lattice disorder) and composition of shallow layers with sub-nm resolution near the surface. This depth resolution is achieved through a careful selection of the experimental conditions that include an ion beam energy for which the inelastic energy loss rates are near the maximum of the stopping power curve, the use of a high energy resolution electrostatic analyser (<0.4%) and the application of a double alignment scattering geometry that maximizes the pathlengths of the particles within the solid. For the case of shallow implanted layers, MEIS has the unique capability of simultaneously providing quantitative, depth distributions of implant damage (i.e. displaced Si atoms) and of most implanted species. Applications of MEIS will be presented first in a study of the evolution of disorder of ultra shallow B and As implants into Si that reveals effects such as dynamic annealing operating at room temperature, disorder accumulation at the surface, epitaxial regrowth and As segregation upon annealing. In these comparatively straightforward, shallow implanted Si samples, quantification can be achieved through the direct conversion of energy scale of the spectrum into a depth scale using well documented, depth/energy dependent energy loss data and the referencing of the backscattered ion yield to the random level from an amorphized Si sample. Secondly, examples of the application of MEIS to the quantitative depth characterization high-k dielectric gate oxide films of thin layers of a few nanometres thickness, will be given. Such ultra thin multilayer systems are pushing the technique to the limit of its capability. The interpretation of MEIS spectra of such more complex multi-component films, becomes increasingly contingent on the availability of a suitable simulation model that takes into account all ballistic effects that occur during the MEIS analysis. Details of such a model will be very briefly outlined and its ability to simulate and interpret the MEIS spectra will be demonstrated in a systematic study of a series of ALD grown HfSiON multilayers of a few nm thickness and the capabilities and limitations of MEIS in terms of depth resolution and quantification discussed.

L2-S2.5
Implantation Induced Amorphization and Graphitization of Single-Crystal Diamond. (#1253) Kevin S Jones¹, D P Hickey¹, Robert G. Elliman²; ¹Department of Materials Science and Engineering, University of Florida, Gainesville, USA ; ²Research School of Physical Sciences and Engineering, Australia.

The amorphization of single-crystal diamond by ion implantation was investigated. A high temperature high pressure process was used to grow flawless single crystal diamonds that were sliced into 5mm diameter (100) wafers. A 1 MeV, 7 x 1015 Si+ cm-2 implant created an amorphous carbon layer ~570 nm thick. The amorphization process was examined using cross-sectional transmission electron microscopy (XTEM). In order to quantify the volume change associated with amorphization a higher energy non-amorphizing implant was used to create a buried marker layer. This layer provided a reference frame to study the swelling from the shallower amorphization. This as-implanted amorphous layer was found to be 25% less dense than crystalline diamond but still 15% more dense than graphite. Annealing between 1 and 24 hrs at 1350 °C, resulted in conversion of the amorphous carbon layer to crystalline graphite. The graphite showed an orientation relationship to the substrate diamond with the (002)graphite parallel to the (022)diamond. Unlike other group IV semiconductors Si and Ge, no evidence of solid phase epitaxial crystallization of the amorphous C was observed upon annealing.