SYMPOSIUM N

9:30 AM *N4-S1.1 (invited)
Heterostructures Inside Nanowires and Nanotrees - Possible and Impossible Combinations. (#1328) Reine Wallenberg, nCHREM, Polymer and Materials Chemistry, Chemistry Center, Lund University, Lund, Sweden.

L. Reine Wallenberg*, Kimberly A. Dick***, Philippe Caroff***, Suneel Kodambaka*, Mark C. Reuter*, Jakob B Wagner* , Knut Deppert,*** Lars Samuelson***, Frances M. Ross** * nCHREM / Polymer & Materials Chemistry, Lund University, Box 124, S-221 00 Lund, Sweden * IBM Research Division, T. J. Watson Research Center, Yorktown Heights, New York 10598 * Solid State Physics, Lund University, Box 118, S-221 00 Lund, Sweden Growing epitaxial semiconductors in a single direction has led to 1-D nanowire devices in III-V materials [1]. A natural step forward is to form branching of the nanowire to enable novel electronic materials with a 3-D structure. We will show structures formed by sequential seeding and growth of primary trunk nanowires, secondary branch nanowires and tertiary leave nanowire assemblies, forming so called nanotrees [2]. The composition can be varied between the levels of branching, and even within each level, to form e.g. LED:s, or resonant tunneling diodes, SET:s and single electron storage devices. The growth method is MOCVD or CBE, and the seeding is normally performed using aerosol-deposition of size-selected Au nanoparticles. Combining seed particle catalysed growth with non-catalysed growth, changes in radial composition can be obtained, to enhance optical properties of light-emitting wire segments. We observe that nanowire heterostructures tend to continue in the same crystallographic direction as the first-grown part when growing one material on top of another (for example, GaP on InP), but to kink into another crystallographic direction when growing the inverted combination (InP on GaP). The first explanation that comes to mind is that the cause is the relative mismatch between the two materials, and for nanowire superlattices, a 3 % mismatch has been reported as acceptable [3]. However, this does not seem to be the ruling factor, since we can grow with comparatively large both positive and negative mismatch. New models based on the supersaturation conditon have emerged.

10:00 AM N4-S1.2
Solving Complex Mesostructures Using Electron Tomography. (#566) Pei Yuan¹, Cheng Zhong Yu¹, Jin Zou²; ¹Fudan University, Shanghai, China ; ²The University of Queensland, Brisbane, Australia.

Recently, helical and circular mesostructures have attracted extensive attention due to their scientific importance and potential applications in chiral separation. To date, the helical mesostructures have been well characterized. For the circular mesostructures, however, it lacks a convincing characterization method to differentiate a close helical (CH) mesostructure with small pitch from a concentric circular (CC) mesostructure. Either CH or CC mesostructures cannot be solely described by the traditional crystallography such as space groups and symmetry elements, thus cannot be determined by the X-ray diffraction or electron diffraction methods. Although conventional transmission electron microscopy (TEM) techniques have been widely used to directly determine various mesostructures, it is still impossible to determine the CH or CC mesostructures with the conventional TEM. This is mainly because the thickness of TEM specimens are generally much larger than the distance between the neighboring pores, so that the TEM contrast caused by the overlapped complex pore structures cannot be used to solve the structure. In this presentation, we will demonstrate that, by using the electron tomography, the complex CC hexagonal siliceous mesostructures can be successfully differentiated from their CH counterpart. MCM-41 type silicas with a hexagonal circular mesostructure and a rod-like morphology were synthesized under basic conditions by using octadecyltrimethyl ammonium bromide (C18TAB) as a template. The distance between two adjacent pores can be measured to be 4.6 nm by TEM. For a tpical rod, a series of tilted TEM images were digitally acquired along two orthogonal axes. To further solve the circular pore architectures, ET and the associated 3D structural reconstruction were carried out. It is of importance to note that the thickness of each artificial slice is only 0.26 nm in our processing, thus even for a helix with the smallest pitch (~ 4.6 nm), ~ 18 slices can be obtained within one pitch, which can reflect the "local" information of helical or circular conformation, thus provide sufficient evidence to determine the real mesostructures. By comparing our results and the simulated model, the possibility of the CH structure can be exclusively ruled out. In conclusion, the true internal structure of the concentric circular hexagonal mesostructure has been solved for the first time using electron tomography technique.

MORNING BREAK 10:30 AM - 11:00 AM

11:00 AM *N4-S2.1 (invited)
Band-Structure Information on the Nanoscale by Valence Electron Energy-Loss Spectroscopy in (S)TEM. (#199) Rolf Erni, National Center for Electron Microscopy, Lawrence Berkeley National Laboratory, Berkeley, California, USA.

The scaling down of semiconductor architectures used for instance in LEDs or the selective utilization of individual quantum dots are examples demonstrating the need for experimental techniques that provide precise band structure information at high spatial resolution. Although standard optical techniques can give superior spectral information, these techniques often lack the spatial resolution to probe materials on the nanometer scale. Valence electron energy-loss spectroscopy (VEELS) in (scanning) transmission electron microscopy (STEM) provides the spatial resolution to locally analyze the electronic structure of heterogeneous nanomaterials and to measure, e.g., bandgaps of individual nanoparticles [1,2]. In particular, it is the utilization of a gun monochromator combined with an illumination aberration corrector that has led to a revival of efforts to exploit the low-loss region of electron energy-loss spectra. The monochromator enhances the energy resolution and thus increases the signal-to-background ratio close to the zero-loss peak where valuable band structure information of semiconductors is contained. The combination of illumination aberration corrector and monochromator allows for adjusting electron probe current, its size and energy resolution over a large range. The experimental conditions, including microscope high tension, can thus be tailored according to the requirements of the material under investigation. Band structure information measured by VEELS can thus directly be correlated with the structural information obtained via STEM imaging. Although numerous examples in the literature demonstrate that VEELS is a reliable technique to identify band structure information of (nano-)materials, the information content of VEEL spectra is controversial. Surface, finite sample size and retardation contribution can alter the dielectric losses measured by VEELS [3]. Dielectric theory [4,5] however provides a mean to critically check experimental data and, provided that (optical) dielectric data are available, to forecast spurious effects in VEELS. In this contribution, several VEELS applications will be presented with focus on studies of electronically and optically active (nano-)materials and devices. In addition, low-loss contributions that can impact the bulk dielectric signal, such as energy losses caused by surface and interface plasmons as well as spectral contributions caused by Cerenkov losses and by energy losses that are due to the excitation of guided light modes, will be discussed. [1] R. Erni, N. D. Browning, Ultramicroscopy 104 (2005) 176. [2] R. Erni, N. D. Browning, Ultramicroscopy 107 (2007) 267. [3] R. Erni, N. D. Browning, Ultramicroscopy 108 (2008) 84. [4] E. Kr?ger, Z. Phys. 216 (1968) 115. [5] J. P. R. Bolton, M. Chen, Ultramicroscopy 60 (1995) 247.

11:30 AM *N4-S2.2 (invited)
New Method for Quantitative Strain Mapping at the Nanoscale in Electronic Devices. (#1317) Martin Hytch, Florent Houdellier, Florian Hue, Etienne Snoeck; Center for Material Elaboration and Structural Studies, National Center for Scientific Research (CEMES/CNRS), Toulouse, France.

Strained silicon is now an integral feature of the latest generation of transistors and electronic devices because of the associated enhancement in carrier mobility. Different ways have been employed to engineer strain in devices leading to complex strain distributions in 2 and 3 dimensions. Developing methods of strain measurement at the nanoscale has therefore been a major goal of recent years but has proven illusive in practice. None of the techniques combine the necessary spatial resolution, precision and field of view. For example, Raman spectroscopy or X-ray diffraction techniques can map strain at the micron scale over wide areas, whilst transmission electron microscopy (TEM) can map strain at the nanometre scale but only over small areas. Here we present a new technique capable of bridging this gap and measuring strain to high precision, with nanometre spatial resolution and for micron fields of view. The method combines the advantages of the conventional moir? technique with the flexibility of off-axis electron holography and is applicable to standard focused-ion beam (FIB) prepared samples. We will present measurements of strain in the active regions of strained-silicon MOSFET devices and compare the results from finite element modelling.

12:00 PM N4-S2.3
Electron Microscopy Observation of Defects in InGaN/GaN Multiple Quantum Wells and AlGaN/GaN Strain-Layer Superlattices. (#316) Jer-Ren Yang, Wei-Chi Li, Hung-Wei Yen; Department of Materials Science and Engineering, National Taiwan University, Taipei, Taiwan.

Devices based on InGaN/GaN quantum wells (QWs) have been widely manufactured as light emitting diodes (LEDs) and laser diodes (LDs). They exhibit the exceptionally high photoemission efficiency despite quite a few threading dislocations (TDs) in InGaN/GaN QWs. This work deals with defects in InGaN/GaN QWs that might be a key of the high photoemission. The high-angle annular dark-field (HAADF) scanning-transmission electron microscopy (STEM) observation manifested the structure of V-shape defects, which comprises thin InGaN/GaN QWs on the {1 0 -1 1} sidewalls. It is proposed the formation mechanism of the V-defects, taking into account the growth kinetics of the GaN crystal and the masking effect of In atoms segregated in the Cottrell atmosphere around the core of TD. It has been believed that there are In-rich regions in the main QWs (on the c plane) acting as QDs, which were found in HAADF-STEM images. The nanostructure of AlGaN/GaN strained-layer superlattice (SLS) cladding in a GaN-based violet laser diode (LD) has also been investigated by HAADF-STEM. It was found that the threading dislocations disappeared within the SLS; this evidence plainly showed the role of SLS in suppressing threading dislocation propagation.

12:15 PM N4-S2.4
High-K HfAlO Charge Trapping Layers for Nano-Scale Non-Volatile Memory Device Applications. (#753) Ting-Yu Wang¹, Zongwen Liu¹, Rongkun Zheng¹, Siddheswar Maikap², Simon Peter Ringer¹; ¹Australian Key Centre for Microscopy & Microanalysis, Australia ; ²Cheng Gung University, Taiwan.

The memory performance of high-k p-doped Si/SiO2/HfAlO/Al2O3/Pt nanolaminate structure, where the high-k Al2O3 and HfAlO charge trapping layers were deposited by atomic layer deposition were investigated and demonstrated to be superior to that for pure HfO2 and pure Al2O3 charging trapping layers. Most significantly, this material demonstrated well-behaved counter clockwise capacitance-voltage hysteresis characteristics for all memory capacitors. A memory window of ~8.6 V with a sweeping voltage of ± 16 V, a high charging trap density of ~1.2 x 10^13 /cm^2 and a large memory window of ~5.7 V after 10 years of retention were observed for the HfAlO charge trapping layer. We have investigated the origins of this performance using scanning transmission electron microscopy and atom probe tomography. Our results indicate the presence of HfAlO nanocrystals, which have been characterised in detail. We propose that these nanocrystallites play an important role in the enhanced memory performance of this structure.

LUNCH 12:30 PM - 2:00 PM

2:00 PM *N4-S3.1 (invited)
Nanoscale Metrology with the Scanning Electron Microscope - Problems, Errors, and Solutions. (#1257) David C Joy, Departments of Materials Science and Engineering, University of Tennessee, Knoxville, USA.

Metrology based on the analysis of images obtained from the scanning electron microscope (SEM) has become of increasing importance over the last three decades as 'critical dimension' scales have been reduced from micrometers to nanometers, and it is still the most widely applied technique for metrology. However the fact that an object can be visualized in the SEM does not necessarily imply that its size and shape can be accurately determined from the image. This is especially true at the nanometer scale level where factors that are of little significance for larger structures become major sources of error. Examples of some of the most significant of these, and steps towards their elimination, are: (1) Unlike an optical microscope, where the image magnification is fixed and well defined, the field of view of an SEM image is calculated as required on the basis of assumptions about the beam energy, the working distance of the sample, and the current flowing through the scan coils. While these quantities are suitably constant for a given instrument, two examples of the same machine would likely generate significantly different estimates of the size of the same feature even under nominally identical conditions. An urgent requirement is there the development of traceable calibration artifacts suitable for the size range between one nanometer and one micrometer. Progress towards the production of SEM length standards will be presented. (2) The scan rasters which deflect the beam across the sample in the SEM are non-linear by amounts which depend on the scan speed and the magnitude of the scan current so in consequence the size of objects in the image will vary depending on where in the field of view they appear. Because every instrument has its own characteristics these errors - which are typically 3-5% or higher must be quantified and then removed numerically. A procedure for the rapid determination of screen non-linearity will be discussed. (3) A majority of the samples to which SEM metrology is applied are semiconducting, or insulating, and therefore can. and usually will, charge under the electron beam. In turn the potentials associated with charging result in dynamic local fluctuations in the imaging magnification which can degrade the precision of a measurement. Experimental data on the magnitude and correction of this effect will be presented. (4) SEM metrology of features in the low-nanometer size range is ultimately constrained by factors such as the penetration of the electron beam into the sample and the three dimensional shape of the electron probe. It will be shown that these factors can be studied, and accounted for, by determining the modulation transfer function (MTF) of the SEM under various imaging conditions. Crucially, variations in the MTF resulting from errors in focusing or in the choice of beam convergence angle effectively result in a distortion of image features. An important recent development is the use of scanning microscopes using Helium ions (SIM) rather than electrons. It will be demonstrated that this development potentially avoids some of the issues discussed for electron beam tools, and the comparative performance of a SIM and a SEM for metrology will be compared.

2:30 PM N4-S3.2
Development of Nanometrology in Australia - Issues Relating to Electron Microscopy. (#895) Asa Katarina Jamting, John Miles; National Measurement Institute, Sydney, New South Wales, Australia.

Nanotechnology is a global phenomenon that is already impacting significantly on Australia's scientific, technological, economic and social development. As nanotechnology develops further in Australia as an industry, it will depend on the provision of a suitable scientific, commercial and regulatory environment. A fundamental element of this environment is measurement. Metrology is the science of measurement and reliable measurements of physical, chemical and biological quantities are required at all stages of the nanotechnology value chain to truly understand and control the manufacturing process and ensure and demonstrate product quality. One of the great challenges with nanotechnology is its ability to span all disciplines of science. In a recent report by the Nanotechnology Taskforce from Department of Industry, Tourism and Resources[1] several areas of interest were identified including health, safety and environment issues; improvement of community awareness and public engagement; development of metrology and standards; coordination of activities amongst federal and state/territory governments; need for international cooperation; industry infrastructure and development. A report from 2006 into Australia's nanometrology needs found that nanotechnology is already relatively strong in the research sector and is beginning to emerge in the business sector[2]. Internationally, a survey carried out in 2005 by the Nanotechnology Research Center (NTRC) at the Industrial Technology Research Institute (ITRI) of Chinese Taipei in the Asia-Pacific region found that some of the areas where accurate measurements on the nanoscale are desired include particle size and thin film thickness and composition[3]. The main tools for these measurements include electron microscopes, both transmission and scanning, as well as scanning probe microscopes. This article outlines the current status of the work in nanometrology at the NMI and aims to highlight some issues relating specifically to the scanning electron microscopy field.

2:45 PM *N4-S3.3 (invited)
Cathodoluminescence of Semiconductor Oxide Nano- and Microwires in the Scanning Electron Microscope. (#116) Javier Piqueras¹, Mariana Chioncel², Ana Cremades¹, Carlos Diaz-Guerra¹, Paloma Fernández¹, David Maestre¹, Yanicet Ortega¹; ¹Department Física de Materiales, Facultad de Ciencias Físicas, Universidad Complutense de Madrid, Spain ; ²Physics Department, Faculty of Chemistry, University of Bucharest, Romania.

Optical, electronic and structural characterization of semiconductor nanowires is a subject of increasing interest due to the potential use of these structures in future nanoelectronic systems, optical nanodevices and other applications. It is known that one-dimensional semiconductor structures have often different electronic behavior than the bulk material, due to factors such as the influence of surface states, defect structure or size effects. For this reason, cathodoluminescence (CL) in SEM, which provides spatially resolved information on radiative recombinations, is a useful complementary technique to characterize semiconducting nanostructures. In this work, CL in SEM has been applied to study nano- and microwires, as well as other low dimensional structures, of the semiconducting oxides,V2O5, Fe2O3, ITO (indium-tin-oxide) and ZnO. The samples were prepared by compacting high purity powder to form disks of about 7 mm diameter, and then annealing under argon flow. This method leads to the growth of elongated structures directly on the sample surface, which acts as source as well as substrate [1] [2]. The growth mechanism is a vapour-solid process which does not involve a foreign substrate or a catalyst. The morphology and luminescence of the obtained nano- and microstructures were investigated by SE imaging and CL in SEM respectively. The structures were also characterized by XRD, TEM and EDS. The formation of the wires induces, in all investigated cases, changes in the CL intensity and/or spectra, as compared with the bulk material. Near band-gap luminescence involving valence band and two conduction bands was found to be favoured in the V2O5 nanowires which also showed a band at 1.70 eV attributed to oxygen vacancies. Fe2O3 nanowires have also been found to show a relatively high CL signal as compared with Fe2O3 powder. CL intensity of the nanowires as a function of temperature shows some anomalies in the range 180K-210 K, whose relation with changes in the magnetic structure of the oxides are currently investigated. CL imaging enabled to characterize ITO nanostructures prepared from mixtures of SnO2 and In2O3 powders. Sn-rich structures show higher emission than In-rich regions. These regions appear frequently as nanopyramids or nanoislands grown on larger Sn-rich microtubes. The incorporation of Sn in doped ZnO nanowires and nanoplates was revealed by local CL spectra which showed a blue shift of the near band gap emission and the overshadow of the defect green luminescence band by a Sn related emission in the range 1.75 eV-2 eV. The CL effects observed in the nano- and microstructures of the above mentioned oxides will be discussed. [1] D.A.Magdas, A.Cremades and J.Piqueras, Appl. Phys. Lett. 88, 113107 (2006) [2] P.Hidalgo, B.M?ndez and J.Piqueras, Nanotechnology 18, 155203 (2007) 1 Permanent address: Physics Dpt., Faculty of Chemistry, University of Bucharest, Romania

3:15 PM N4-S3.4
Probing the Cathodoluminescence Properties of the ZnO:Zn Surface. (#1261) Matthew R. Phillips¹, Cuong Ton-That¹, Dominique Drouin²; ¹University of Technology Sydney, Sydney, New South Wales, Australia ; ²Université de Sherbrooke, Québec, Canada.

ZnO:Zn phosphor powder with a mean diameter of 1 micron has been studied using ultra-low voltage (200V) scanning cathodoluminescence microscopy and spectroscopy. Using the CASINO Monte Carlo simulation it can be shown that at 200V the electron range in ZnO is < 2nm and the post-diffusion electron-hole pair distribution is primarily located within the top 20nm of the sample. The cathodoluminescence (CL) emission at 200V was found to consist of strong green emission centred at 2.50 eV with a weak near band edge peak at 3.37 eV. Gaussian peak fitting revealed that the broad green emission consisted of two emission peaks centred at 2.50 eV and 2.26 eV. Excitation density CL analysis (ICL ~ IB^n) was conducted over a beam current (IB) range of 1 to 1000 pA. Plotting logICL versus logIB revealed a linear relationship for IB > 100 pA providing a power law exponent of n = 1 and n = 0.5 for the 2.50 eV and 2.21 eV peaks respectively. These data indicate that the recombination centres responsible for the 2.50 eV emission (n = 1) are abundant and have a fast relaxation time. However, the sub-linear power law exponent (n = 0.5) for the 2.21 eV peak indicates saturation of the emission with increasing IB, suggesting recombination at a limited number of defect centres and/or slow relaxation times. Recent theoretical studies of the surface electronic structure of the ZnO surface will discussed in context of these experimental results.

AFTERNOON BREAK 3:30 PM - 4:00 PM

4:00 PM *N4-S4.1 (invited)
Investigation of the Microscopic Local Recombination Dynamics in A-Plane GaN ELO Structures by Spectrally and ps-Time-Resolved Cathodoluminescence Microscopy. (#920) Barbara Bastek¹, Frank Bertram¹, Juergen Christen¹, Tim Wernicke², Markus Weyers², Michael Kneissl²; ¹Otto von Guericke University of Magdeburg, Germany ; ²Ferdinand-Braun-Insitut für Höchstfrequenztechnik Berlin.

A promising approach to overcome the problems of quantum confined Stark effect is the growth of non-polar a-plane GaN. However, it still suffers from non-mature morphological quality of the epilayers. The technique of epitaxial lateral overgrowth (ELO) has been proven very effective for c-axis grown GaN. We report on ELO of a-plane GaN applied to improve the crystallographic material quality. Fully coalesced a-plane GaN layers were grown by MOVPE on r-plane sapphire substrate using stripe masks aligned along the [01-10] direction. The individual characteristic ELO domains were directly imaged by highly spatially and spectrally resolved cathodoluminescence microscopy (CL) at 5K: In the area of coherent growth (I), i.e. directly above the mask openings, the broad luminescence band of the basal plane stacking fault (BSF) centered at 3.41 eV dominates the emission. In the region of coalescence (II) the BSF luminescence dominates as well. In complete contrast, in the stripes associated with the laterally overgrown domains in [0001]-direction (III), exclusively an intense and sharp (D0,X) emission at 3.475 eV is observed. ps-time-resolved CL was performed locally on the individual growth domains at different spectral positions. The CL of the BSF shows a strictly non-exponential behavior, i.e. a slowing down of decay with increasing delay time. The initial lifetime of the BSF luminescence, i.e. ?(t&rarr0) exponentially decreases with rising temperature yielding a thermal activation energy of 7 meV. This value nicely fits the valence band discontinuity obtained from the Arrhenius behavior of the BSF luminescence intensity. In complete contrast the temperature dependence of the FX initial lifetime recorded from the domain (III) shows a non monotonous behavior. The initial lifetime of 180 ps at 5K is dominated by the fast capture of FX by donors forming (D0,X) plus the capture into the quantum well like BSFs. With rising temperature this capture time constant decreases as T^(-1/4) and reaches a minimum of 104 ps at T=60K. Above this temperature, i.e. when FX starts dominating the spectrum, the FX lifetime rapidly increases reaching a value of 240 ps for 300K. In parallel we observe a drastic change in the BSF/NBG intensity ratio for temperatures from 50 to 100K. While for low temperatures the BSF recombination dominates the overall emission for higher temperatures the FX emission clearly dominates the spectrum. This behavior can be interpreted as the emission of holes from the BSF-QW for temperatures above 60K. The fully spectrally resolved recombination dynamics shows a strong monotonous red shift of the BSF luminescence of 11 meV within 22 ns after end of excitation. In complete contrast the (D0,X) luminescence stays spectrally at the same position. This might be an evidence for the reduction of the screened quantum confined Stark effect in the BSF quantum wells with decreasing carrier density with progressive decay time

4:30 PM *N4-S4.2 (invited)
Low-Voltage Cathodoluminescence Microscopy and Spectroscopy of Diluted InAs/InP Self-Assembled Quantum Dots. (#1262) Emmanuel Dupuy¹, Nicolas Pauc¹, Vincent Aimez¹, Michel Gendry², Denis Morris¹, Dominique Drouin¹; ¹Departement de Génie Électrique et en Génie Informatique, Université de Sherbrooke, Québec, Canada ; ²Institut des Nanotechnologies de Lyon, France.

The development of a wide variety of new materials and devices to be used in nanoelectronic and photonic applications has required powerful tools for the characterization of both bulk and microstructure samples. The Scanning Electron Microscope (SEM) combined with Cathodoluminescence plays a key role in the characterization of a wide variety of these materials. Cathodoluminescence (CL) resolution is limited by two processes which are the primary electrons scattering and diffusion of generated carriers before capture and recombination. Using a low-voltage acceleration source for the primary electrons, the interaction volume into which electron-hole pairs are created is reduced to a few nanometres. Under these conditions, the CL resolution is essentially limited by the diffusion length of the generated carriers in the medium. Consequently, carrier behaviour can be investigated using scanning cathodoluminescence microscopy and spectroscopy at high spatial resolution. In this contribution, we present measurements of carrier-diffusion lengths on self-assembled quantum dots (QDs) by low-voltage cathodoluminescence. Measurements were performed on diluted InAs QDs grown on InP(001) substrate by molecular beam epitaxy. These QDs are promising active structures for the realization of optoelectronic devices operating at 1.55 um, such as single-photon sources for quantum cryptography systems. Specific growth parameters were used to produce diluted and coherent InAs/InP(001) QDs whose density can be reduced in the 1 um-2 range. QDs are capped with a thin 10-20nm InP thickness layer to collect enough CL signal and emit near 1.5 um at room temperature. At 1 kV, we spatially resolve individual QDs which appear as correlated bright spots. Analysis of CL intensity profiles across single QDs provide a direct measurement of the characteristic carrier-diffusion length in the InP barrier before capture. Temperature dependence will be discussed.

5:00 PM N4-S4.3
The Affect of the Purity of Reactants and Post-Processing Techniques on the Cathodoluminescence of ZnO Nano-Powders. (#991) Katie Elise McBean¹, Matthew Ronald Phillips¹, Dominique Drouin²; ¹University of Technology Sydney, Sydney, New South Wales, Australia ; ²Université de Sherbrooke, Québec, Canada.

ZnO nano-powders were synthesised by mixing equal volumes of a 0.05 M zinc chloride ethanolic solution with 0.20 M sodium hydroxide (either 97+% or semiconductor grade) ethanolic solution, for six hours at room temperature. These cloudy solutions were centrifuged and the powders rinsed several times before being dried under vacuum. These ZnO powders are identified as: ZnO; made with 97+% NaOH or HP ZnO; made with the semiconductor grade NaOH. Cathodoluminescence (CL) spectra and images of the as-prepared specimens were collected at 5 K, in addition to specimens annealed at 700 C for 10 minutes in an atmosphere of 5% Hydrogen. The HP ZnO displayed a much more intense UV near band edge (NBE) emission compared with ZnO, and lacked the broad green defect related emission observed in the specimen prepared with 97+% NaOH. Annealing in a Hydrogen atmosphere increased the intensity of the NBE emission in both powders although the HP ZnO NBE emission remained by far the most intense, while the intensity of the green defect emission of the ZnO was greatly reduced. The NBE emission from Hydrogen annealed HP ZnO contained recombination centres at 3.36 eV; attributed to I6/6a Aluminium donor and 3.32 eV; an unidentified acceptor-bound exciton emission with LO phonon replicas at 3.25 eV and 3.18 eV. The Hydrogen annealed ZnO displayed weaker intensity emissions at the same energies in addition to a weak green defect emission at 2.40 eV.

N4-S5.1
Effect of Titanium on the Nanostructure of Centerline Precipitates in Low Carbon, Low Manganese Steel Slabs. (#1190) Sima Aminorroaya-Yamini, Rian Dippenaar; The University of Wollongong, New South Wales, Australia.

Elongated inclusions, particularly MnS, contribute significantly to reduced ductility and toughness in hot rolled steel but earlier research indicated that these properties can be improved by titanium additions. Such addition to a steel result in titanium being dissolved in manganese sulphide or MnS being replaced by TiS and/or titanium carbosulphides. In the present study a steel was designed to decrease alloying element segregation and to evaluate the effect of titanium on centerline sulphide precipitates. Precipitates were identified by using scanning electrom microscopy and characterised by the use of transmission electron microscopy following sample preparation by focused ion beam milling techniques. Iron-titanium-sulphides and MnS precipitates containing iron, form in close proximity to each other. Evidence is provided that an increase in the titanium content of steel leads to an increase in the percentage of titanium contained in the iron sulphides and a decrease in the iron content of MnS inclusions.

N4-S5.2
The Australian Nanotechnology Toolkit: Advanced Microscopy and Microanalysis. (#555) Miles Apperley¹, Simon Peter Ringer², Paul Munroe³, John Drennan⁴, Tim Senden⁵, Hans Greisser⁶, Joe Shapter⁷, John Terlet⁸, David Sampson⁹; ¹Australian Microscopy and Microanalysis Research Facility, Australia ; ²The University of Sydney, Australia ; ³University of New South Wales, Australia ; ⁴The University of Queensland, Australia ; ⁵Australian National University, Australia ; ⁶University of South Australia, Australia ; ⁷Flinders University, Australia ; ⁸The University of Adelaide, Australia ; ⁹University of Wetern Australia, Australia.

Australian nanotechnology research and development, and ultimately the capacity to be innovative in these areas, requires a broad base of national resources. These resources must include access to world-class instrumentation and expertise at research facilities for the characterisation of materials and biological systems down to the nanomolecular length scale. These capabilities are essential if nanoscale fabrication or molecular processes are to be understood and developed into new business opportunities for Australia. The Australian Microscopy and Microanalysis Research Facility (AMMRF) provides the peak Australian facility for the characterisation of materials through macro, meso, nano and atomic length scales by means of advanced microscopy and microanalysis. It has been formed as a joint venture between the universities of Sydney, Queensland, New South Wales, Western Australia, Australian National University and the South Australian Regional Facility - a collaboration between The University of Adelaide, The University of South Australia and Flinders University. At the core of the AMMRF is the transformation of stand-alone microscopy and microanalysis centres at these universities into a single national network of laboratories, unified in terms of both equipment and research expertise. Into this co-laboratory a top tier or suite of flagship instruments were added that represent the leading edge in nanostructural analysis in the world. These are tools are available to all Australian scientists at research institutions on a merit basis and nominal fees. The advanced facilities are also available to industry under suitable commercial arrangements. Importantly, AMMRF is more than a collection of characterisation instruments. It is the fusion of advanced infrastructure and technical expertise with in-house research programs that enables users of the facilities to draw on a comprehensive resource. Research outcomes arise from the common 'user experience' that access to AMMRF entails. This experience comprises the steps of formulating the initial idea or problem, planning a project, undertaking training, using equipment and collecting data, interpreting the data and solving problems, and finally using and presenting the outcomes in a variety of formats. The toolkit of advanced microscopy and microanalysis capability that is available to Australian material science and nanotechnology researchers will be reviewed in this paper. The power of techniques such as atom probe tomography, high resolution scanning electron microscopy coupled with focussed ion beam milling, transmission electron microscopy, state-of the art scanning probe microscopy and nanoscale mass spectroscopy, to enable world class science outcomes will be showcased.

N4-S5.3
Tin Whiskers on Lead-Free Printed Circuit Board after Finishing. (#1092) Ulrich Beck, Mathias Nowottnick; Institute of Electronic Appliances and Circuits, Faculty of Computer Science and Electrical Engineering, University of Rostock, Germany.

In the 1960th and 1970th while use of plumbiferous solders the whisker growth on circuits and on finished printed circuit boards was a well-known problem. In the following years these problems - the whisker growth - could be eliminated. For a short time we detected newly Tin whisker growth on lead-free printed circuit boards of various manufacturers if theses were assembled with pressed in contact pins. We analyse with REM and EDX the whiskers origin, whiskers growth and the sources of the whiskers. Because of the importance of the whisker growth for electronic devices we will give some proposals to avoid whisker growth.

N4-S5.4
Influence of Silicon Dioxide Barrier Layer on the Microstructure of PSZT Thin Films on Gold-Coated Silicon. (#130) Madhu Bhaskaran¹, Sharath Sriram¹, David R. G. Mitchell², Anthony S. Holland¹, Arnan Mitchell¹; ¹RMIT University, Melbourne, Victoria, Australia ; ²Australian Nuclear Science and Technology Organisation, Australia.

Strontium-doped lead zirconate titanate (PSZT) belongs to the family of doped lead zirconate titanate (PZT) compounds, and is reputed to exhibit relatively high levels of piezoelectric response. To maximise the level of piezoelectric response from PSZT thin films, preferentially oriented perovskite-structured films are desired. This article reports on the influence of the gold bottom electrode and the presence or absence of an intermediate silicon dioxide layer on the orientation of deposited PSZT thin films. Gold was chosen due to the proximity of its lattice spacings to those of PSZT. Cross-sectional transmission electron microscopy (XTEM) analysis of samples with PSZT thin films deposited at 300 °C showed evidence of gold reacting with silicon to form faceted crystallites projecting into the substrate. The gold-silicon reaction also resulted in an amorphous silicon dioxide layer on the surface of gold, as a result of the fast outward diffusion by silicon through grain boundaries in gold. The formation of this amorphous layer during the subsequent initial stages of PSZT deposition prevented the gold layer from influencing the structure of PSZT. To prevent this gold-silicon reaction, an intermediate silicon dioxide layer was introduced between the metal layers and silicon. The PSZT deposition temperature initiates crystal growth in the gold layer (which was nanocrystalline when deposited), resulting in significant changes to the texture of gold. X-ray diffraction and XTEM studies indicate that there is an increase in preferential orientation in the gold layer by a factor of 6 (compared to a sample without the silicon dioxide layer). In the absence of gold-silicon interactions, the gold crystallites are able to reorient so that the most densely packed planes (111) are parallel to the surface, and thus the surface energy is minimised. The formation of the amorphous layer is also prevented. As a result, the highly textured gold layer was able exert a strong guiding effect on the subsequently deposited PSZT thin films, which developed a pronounced (104) texture. The increased preferential orientation of these PSZT thin films resulted in the films exhibiting ultra-high piezoelectric response (50% greater than films without the intermediate silicon dioxide layer).

N4-S5.5
Electron Microscopic Investigation of Carbon and Boron Nitride Encapsulated Nanoparticles Prepared by High Energy Ball Milling and Arc Discharge. (#1297) Boris Borisovich Bokhonov, Marat Rashidovich Sharafutdinov; Institute of Solid State Chemistry and Mechanochemistry, Siberian Branch of Russian Academy of Science, Novosibirsk, Russian Federation.

The discovery of carbon nanotubes, encapsulated nanoparticles has stimulated a great interest in the study of filled nanostructures and promotes searches for new methods of preparation of nanocomposites. Many metals, carbides, nitride, boride and oxides have been encapsulated in carbon or boron nitride nanocapsules by various methods, e.g., chemical methods, arc discharge, arc melting, catalyzed hydrocarbon pyrolysis . One of the most promising methods to prepare nano-sized composition materials is the mechanical treatment of solids. The ball milling and arc discharge is well known methods for preparation nano-sized materials for various classes of chemical compounds. The goal of the present work was to investigate the changes of structural, phase and morphological characteristics of the mixtures of metal and amorphous carbon or boron nitride during mechanical treatment and annealing. The investigation of morphological and structural changes during high energy ball milling, thermal annealing, arc discharge of the mixtures amorphous carbon/metal (Fe, Co, Ni, Mo, W, Cu, Bi) demonstrated that the activation is accompanied by the formation of nano-sized metal and carbon particles. Application of in situ electron microscopy and x-ray synchrotron radiation allowed us indicate structural and chemical transformation of encapsulated nanoparticles during heating and oxidizing of carbon encapsulated nanorarticles. The annealing of amorphous carbon/metal (Fe, Co, Ni, Mo, W) samples obtained after short-time mechanical activation causes a crystallization of the amorphous carbon as onion-like graphite-metal structures. Annealing of the amorphous carbon/metal samples after prolong mechanical treatment leads to the formation of metal (metal carbide) nanoparticles (40-50nm) encapsulated by graphite. Electron microscopic investigations and X-ray studies of mechanically activated mixture of metal (Fe, Co, Ni, Hf) and boron nitride provide an evidence of the transformation of hexagonal boron nitride into amorphous state. According to the obtained electron microscopic data, the particles of metal and boron nitride after mechanical activation form aggregates has distribution from several nanometer to 100nm. Low temperature annealing of ball milled metal/boron nitride samples caused the formation of nano-sized encapsulated particles of metal boride. The surface of boride nanoparticles were coated with a shell composed of hexagonal boron nitride 5-15 nm thick. The size of encapsulated particles varies within rather broad range from 30 -50 nm to several hundred nanometers. The experimental results allowed us to conclude that the application of mechanical activation with low temperature annealing can be a useful method for large scale preparation of nano-sized encapsulated particles of various compounds (metals, carbide, boride).

N4-S5.7
Cathodoluminescence for High Resolution Non-Destructive Luminescence Depth Profiling. (#998) Annette Dowd, University of Technology Sydney, Sydney, New South Wales, Australia.

We have used cathodoluminescence to study the light emission from Si nanocrystals embedded in SiO₂. Si nanocrystal light emission has previously been studied with photoluminescence however the use of cathodoluminescence allows the possibility of high spatial resolution in 3 dimensions. We have developed a novel non-destructive method to extract the depth distribution of the luminescence intensity using cathodoluminescence. The Si nanocrystal samples were synthesised using high temperature annealing of 600 keV Si⁺ implanted SiO₂ resulting in a band of ~ 5 nm nanocrystals 0.8 um from the SiO₂ surface. Cathodoluminescence was excited with the focused beam of a scanning electron microscope, which was varied in energy from 6 to 26 eV. We observed the emission peak in the near infrared spectral region, centred at approximately 850 nm, in addition to blue and red peaks attributed to SiO₂ defects. We have extended the constant power method described by Fleischer et al [1] by using more advanced data treatment to reveal the distribution of luminescent centres as a function of depth in SiO2. By using the Maximum Entropy method and no a priori assumptions of depth profiles, we obtain a solution of the depth distribution of the 850 nm, with maximum density at a depth of 500±50 nm. This result is explained in terms of inhomogeneous size distribution of Si nanocrystals. The Maximum Entropy method is also used to investigate the depth distribution of the blue and red luminescent centres. [1]K. Fleischer, M. Toth, and M.R. Phillips, Appl. Phys. Lett., 74, 1114 (1999).

N4-S5.8
Nanoscale X-Ray Fluorescence Analysis by Using Planar Waveguide-Resonator. (#344) Vladimir Egorov¹, Evgeniy Egorov¹, Mihail Afanas'ev²; ¹Institute of Microelectronics Technology and High-Purity Materials, Russian Academy of Sciences, IMT RAS, Chernogolovka, Moscow, Russian Federation ; ²MIR&A, Russian Federation.

The problem of X-ray fluorescence analysis application for material nanosize areas rests on difficulties connected with the nanosize zond preparation. It can be solved on base of the focusing zone plate technology, in principle. But this technology is very expensive and fractured. There is more effective to use a simple and cheap technique included the planar X-ray waveguide-resonator (PXWR) [1]. This device even at appearance of a simplest construction is able to form X-ray filiform flux with width 7-200 nanometers and angular divergence 0.1 degree (~1.7*10^-3 rad). The complicated construction utilization, for example, the composite planar X-ray waveguide-resonator, leads to the angular divergence decreasing up to 0.001 degree (~1.7*10^-5 rad) at an integral intensity safety of the emergent flux characteristic component. The premium parameter realization allows to expect that PXWR will be beautiful ground for X-ray fluorescence analysis instruments with nanoscale resolution. At the same time, the most practical efficiency from PXWR application is expected for TXRF spectrometry. This fine analytical method is oriented on the filiform exciting flux using at minimization of it's width and divergence. TXRF exciting flux preparation by PXWR using is more effective then application of the slit-cut systems owing to the significant radiation density increasing in the flux formed by waveguide-resonators. Moreover, the flux formation by PXWR allows to use some independent radiation sources working in the parallel regime. The work introduces to detail discussion of X-ray schemes features used for TXRF spectrometer building. Furthermore, it submits the experimental results presented the analytical efficiency comparison of a conventional TXRF spectrometer in condition of an exciting flux formation by slit-cut system and by waveguide-resonance units. [1] V.K. Egorov, E.V. Egorov. Background of X-ray nanophotonics based on the planar air waveguide-resonators // X-ray Spectrometry. v36. 2007. pp. 381-397. [2] V.K. Egorov, E.V. Egorov. Composite X-ray waveguide-resonator as a background for a new generation of the material testing equipment for films on Si substrate // Processing of MRS. v716. 2007. pp. 189-195.

N4-S5.9
TEM Characterisation of Highly Luminescent CdS Nanocrystals. (#778) Hadas Katz, Alexey Izgorodin, Douglas R. MacFarlane, Joanne Etheridge; Department of Materials Engineering, Monash University, Victoria, Australia.

Luminescent II-VI semiconductor nanocrystals have been the focus of many studies in recent years due to their low energy consumption, wide variety of electroluminescence properties and a large number of combinations of core/shell materials that can be synthesized by the simple and cost efficient reverse micelle method. The size of the nano-crystal determines its surface to volume ratio[1], which affects the band gap and hence the luminescence properties of the crystal. The crystal structure and defect structure, such as point defects and dislocations, also affect band gap energy[2] and hence wavelength of the emitted light as well as the stability of the luminescence materials over time. Characterizing composition, nanocrystal size, atomic structure, defect structure and atomic bonding by electron microscopy will enable us to understand how their size and atomic structure would affect the wavelength of the emitted light and stability of the luminescence materials over time. These are important factors in the engineering of luminescence materials with the desired properties. This work presents an electron microscopy and diffraction study of the crystal structure of highly luminescence CdS nanocrystals produced using the reverse micelles method. Energy-dispersive X-ray spectroscopy (EDX), selected area diffraction (SAD) and atomic resolution imaging using an analytical JEOL 2011 TEM fitted with a LaB6 filament were used to determine the CdS nanocrystal's composition, crystallographic structure, defect structures and size with a view to understanding how these affect the stability, band gap energy and luminescence properties. Nanoparticles were observed both aggregated in clusters and distributed across the carbon film. The size of the nanoparticles is typically between 3-13nm and was determined by counting of atomic planes in atomic resolution images. SAD patterns taken from filtered solution indicate the presence of hexagonal CdS and cubic CdS only. However, in unfiltered solutions cubic CdO and cubic Na2S nanocrystals were also observed and are assumed to be by-products. Careful measurement of a SAD pattern of numerous particles distributed across the carbon film shows a 4% difference in the nominal cubic 200 and 020 spacings, suggesting the cubic CdS structure is distorted. This could have an affect on band gap and electroluminescence properties. Using high resolution images and their Fourier transforms, it was confirmed that both cubic and hexagonal CdS nanocrytals and a cubic CdO by-product are produced using the reverse micelle process. In addition, a high resolution image of a small cluster containing 3 cubic CdS nanocrystals shows that these particles share a common atomic plane and may have inter-grown during growth process. References: [1] S. J. Rosenthal, J. McBride, S. J. Pennycook, L.C. Feldman, Surface Science Reports, 62 (2007) 111-157. [2] R. Xu, Y. Wang, G. Jia, W. Xu, S. Liang, D. Yin, Journal of Crystal Growth, 299 (2007) 28-33.

N4-S5.10
Electrode Erosion and Nanoparticles Reduction in Solution Plasma Process. (#536) Stepan Potocky, Nagahiro Saito, Osamu Takai; Nagoya University, Aichi, Japan.

During recent years plasma systems in liquid have become of topical interest. It is mainly due to their possible application for decomposition (waste water treatment and purification) and on the other hand for production of nanoparticles. These systems are able to produce highly active species which then finally results in conversion of the organic to innocuous materials or in reduction of metal salts. Nevertheless, the drawback of such system meets with the requirement of very high voltage for sustaining the plasma discharge and relatively low energy efficiency. Important process is the erosion of the electrodes. It can limits the operating lifetime but on the other hand it can serve as an important part of nanoparticles reduction process.

We demonstrate operation of solution plasma process under relatively low discharge voltage (below 4 kV) and two different plasma regimes with pulse energies ≈20 mJ using a high frequency bipolar pulsed DC power supply. It can be operated up to the repetition frequency of 30 kHz with a pulse width in the range from 2 μs to 10 μs and a maximum voltage and current of ±6 kV and 7 A, respectively.

Erosion of pin electrodes together with reduction of metal nanoparticles in the pulsed discharge solution plasma process was investigated in dependence on the electrode material and operated plasma regime (two orders difference in hydroxyl radical emission line intensity). The effects on electrode material and discharge regime are discussed. The current-voltage characteristics of the system together with optical emission spectroscopy of plasma discharge were used to characterize different regimes of plasma discharge operation. The morphology of electrodes and produced nanoparticles have been examined by scanning electron microscopy. Metal concentration was analysed using inductively coupled plasma mass spectroscopy to characterize production efficiency. Analysis of generated gas species (hydrogen and oxygen) by using gas chromatography and hydrogen peroxide production by colorimetric method using titanium reagent and the absorption spectroscopy were used to characterize chemical yields.

N4-S5.11
TEM Characterisation of Nanofabrication Process in a Single Crystal Diamond. (#204) Sergey Rubanov, Barbara Fairchild, Paolo Olivero, Andrew Greentree, Steven Prawer; Bio21 Institute, The University of Melbourne, Parkville, Victoria, Australia.

Diamond has attracted enormous interest as a solid state platform for quantum information processing [1-4]. It is highly desirable to fabricate photonic components in diamond on nano-scale level. Recently we demonstrated fabrication of three-dimensional structures in a single crystal diamond. [5]. High fluence MeV ion implantation was used to create a buried damage layer and eventually a graphite-like layer upon annealing. The etchable graphitic layer can be removed to form a free standing membrane into which the desired structures can be sculpted using focused ion beam (FIB) milling. Formation of thin diamond films and buried damage layers was studied using cross-sectional TEM. Single crystal diamond samples were implanted with 2 MeV He+ ions to a fluence of 3-10x1016 ions?cm-2 at room temperature. To create thin diamond layers second implantation with 1.8 MeV He+ was used. In this case two damage layers with a single-crystal layer 'sandwiched' between were created. TEM showed an amorphous layer between two heavily damaged crystalline interfaces. The depth of damage layer correlates well with that calculated using the SRIM2003 program [6]. In case of double implantation TEM showed two buried damage layers with thin crystalline diamond layer between. The sample was thermally annealed at 550o C in air for 1 hour. The annealing results in the formation of a new phase in the middle of the damage layer with thickness ~ 110 nm. The thickness of the damage region remains 180 nm which probably indicates the absence of solid phase recrystallisation of bulk diamond. Selected area diffraction and the HREM imaging of central region showed that the dominant phase is fine grain polycrystalline graphite. However, there are still present two thin amorphous layers between graphitic phase and crystalline diamond interfaces. That means that there is some critical defect density for graphitization at a given annealing temperature and another slightly lower critical density for amorphisation. Higher temperature annealing will be required to remove this residual damage. To create 3-D device structures the samples after ion implantation and thermal annealing were patterned using FIB. The graphitic layers were etched away in boiling acid. The released surface layers were lifted off, leaving behind the desired structures. References [1] A.D. Greentree et al., J. Phys.: Condens. Matter 18 (2006) S825. [2] F. Jelezko et al., Phys. Rev. let. 93 (2004)130501. [3] M.S. Shahriar et al., Phys. Rev. A 66 (2002) 032301. [4] S. Tomljenovic-Hanic et al., Optics Express 14(8) (2006) 3556. [5] P. Olivero et al., Adv. Mater. 17 (2005) 2427. [6] J.F. Ziegler, see: www.SRIM.org.

N4-S5.12
Ultra-High Piezoelectric Response of Strontium-Doped PZT Thin Films Deposited at Low Temperatures. (#132) Sharath Sriram¹, Madhu Bhaskaran¹, David R. G. Mitchell², Arnan Mitchell¹; ¹RMIT University, Melbourne, Victoria, Australia ; ²Australian Nuclear Science and Technology Organisation, Australia.

Applications of piezoelectrics are multi-disciplinary as they have been integrated into transducers, actuators, sensors, drug delivery systems, and nanoscale manipulators, and are potential candidates for use in energy scavenging devices. To improve the efficiency of these applications, high performance piezoelectrics which can be integrated into standard mass fabrication processes are critical. Lead zirconate titanate (PZT), in both bulk and thin forms, is used in most piezoelectric applications due to its high piezoelectric response coefficients. Strontium-doped lead zirconate titanate (PSZT) has shown improved piezoelectric response characteristics. In this work, we discuss the results obtained from the microstructural and piezoelectric response characterisation of PSZT thin films. Thin films deposited on gold-coated silicon substrates were studied using cross-sectional transmission electron microscopy (XTEM) and X-ray diffraction (XRD) techniques. A guiding effect from the gold layer on the PSZT thin films was observed. This resulted in a PSZT rhombohedral unit cell in which 'a' and 'c' were both about 5.29% larger than expected. Crystal structure simulations confirmed the ability of gold to have a guiding effect on this modified unit cell, and these simulations highlight the small lattice mismatches between the gold (111) and PSZT (104) planes. A combination of factors, including the preferential orientation and the modified unit cell, contribute to attaining an ultra-high piezoelectric response. An effective piezoelectric coefficient (d33) value of 892 pm/V was estimated, which is over two times higher than the maximum value of 419 pm/V reported for PZT thin films on silicon. This value of 892 pm/V represents the highest d33 attained for piezoelectric thin films on silicon. An additional consequence of this investigation is the identification of deposition conditions to attain preferentially oriented thin films at 300 °C rather than the conventional 650 °C. These factors (deposition on silicon and lower deposition temperature) makes the deposition process more compatible with multi-layer microsystem fabrication processes, which is highly advantageous for the incorporation of these piezoelectric thin films into multi-disciplinary applications.

N4-S5.13
Modified Unit Cell of Preferentially Oriented Strontium-Doped PZT Thin Films on Pt/TiO₂/Si. (#131) Sharath Sriram¹, Madhu Bhaskaran¹, David R. G. Mitchell², Ken T. Short², Anthony S. Holland¹; ¹RMIT University, Melbourne, Victoria, Australia ; ²Australian Nuclear Science and Technology Organisation, Australia.

The addition of strontium as a dopant to the popular piezoelectric compound lead zirconate titanate (PZT) is reported to enhance the piezoelectric response. Piezoelectric thin films of strontium-doped lead zirconate titanate (PSZT) have been deposited on platinised silicon substrates by RF magnetron sputtering under optimised conditions. These thin films have been studied using a combination of cross-sectional transmission electron microscopy (XTEM), field emission gun scanning electron microscopy (FEG-SEM), selected area electron diffraction (SAED), and glancing angle X-ray diffraction (GA-XRD). GA-XRD peaks highlighted the existence of a crystalline phase unrelated to the calculated powder diffraction file (PDF), with the deposited thin films on Pt/TiO2/Si being preferentially oriented in this phase. Analyses of the films by XTEM highlighted the presence of tightly packed nanoscale columnar grains. Dark field imaging, with and without hollow cone illumination, indicated the existence of two dominant sets of orientations. A combination of the GA-XRD and selected area diffraction results verified the existence of a modified rhombohedral unit cell. This modified cell differs by only 1% in volume from the expected cell, but has a 9.5% larger 'c' and 4.9% smaller 'a'. These results indicate that these strontium-doped PZT thin films might have the capacity to exhibit a higher than expected piezoelectric response, with more room for atomic displacements and larger polarisation bond lengths in this modified unit cell structure.

N4-S5.14
Analysis of Self-Assembled Monolayer Interfaces by Three-Dimensional Atom Probe Tomography. (#240) Wenrong Yang, Rongkun Zheng, Baptiste Gault, Filip Braet, Simon Ringer; The University of Sydney, New South Wales, Australia.

Analysis of Self-Assembled Monolayer Interfaces by Three-Dimensional Atom Probe Tomography Wenrong Yang, Rongkun Zheng, Baptiste Gault, Filip Braet and Simon Ringer Australian Key Centre for Microscopy and Microanalysis, The University of Sydney, NSW, 2006 Australia The molecular-level control over surface modification afforded by alkanethiol self-assembled monolayers (SAMs) makes them attractive as a form of nanotechnology for so called 'bottom up' nanofabrication.[1 ]Characterisation of these modified surfaces presents great challenges. X-ray photoelectron spectroscopy (XPS) is commonly used for surface analysis because it is sensitive to elemental and functional group composition and can be quantified. However, XPS suffers from a lack of chemical resolution and is often unable to determine the overall chemical structures of materials adsorbed on surfaces. Mass spectrometry, on the other hand, is ideal for providing chemical structure, and has been tried for the characterisation of SAM-modified surfaces.[2] Now, whereas mass spectrometry has excellent compositional sensitivity, it does not have the spatial resolution to resolve clearly SAM interfaces. Combining both outstanding chemical sensitivity and atomic resolution, atom probe tomography (APT) is uniquely positioned to provide 3-D imaging and quantitative compositional information on nanostructures.[3,4] Here, we present our recent results on the characterisation of SAM interfaces on metal surfaces using APT and the implications for how this information can feed back into the design of nanofabrication processes. References: [1] Love, J. C.; Estroff, L. A.; Kriebel, J. K.; Nuzzo, R. G.; Whitesides, G. M. Chem Rev 2005, 105, 1103-1169. [2] Yang, W. R.; Zhang, R.; Willett, G. D.; Hibbert, D. B.; Gooding, J. J. Anal. Chem. 2003, 75, 6741-6744. [3] Kelly, T. F.; Miller, M. K. Review of Scientific Instruments 2007, 78, -. [4] Braet, F.; Soon, L.; Kelly, T. F.; Larson, D. J.; Ringer, S. P. Nanotechnology for the Life Sciences 2007, 3, 292-318.

N4-S5.15
Hydrogen Reduction of Titanium Dioxide: SEM & TEM Analysis. (#944) Waven Zhang, Jeffrey C McCallum, David R Sadedin; Department of Chemical & Biomlocular Engineering, The University of Melbourne, Parkville, Victoria, Australia.

The reduction of titanium dioxide by hydrogen plasma is a technical challenge. If successful, it would open the door to an environmentally friendly and economical method to produce titanium. To date, various attempts to reduce TiO₂ using hydrogen plasma have obtained TiO at best, contrary to thermodynamic predictions based on the reactivity of hydrogen atoms. From experiments conducted recently, XRD analysis of the samples indicated that TiO₂ have been reduced to Ti₂O₃ in most circumstances. Analysis by Raman Spectroscopy however, instead indicate reduction to TiH₂. It was suspected that these two analysis methods are sensitive to different depth within the sample. SEM and TEM analysis of the samples have shed light on the discrepancy between these two analysis methods, and the difference in their depth sensitivity.

N4-S5.16
Luminescence Nano-Characterization of Free Standing InGaN/GaN Micro-Disk LEDs on Silicon. (#1071) Alexander Franke¹, Juergen Christen¹, Frank Bertram¹, X.K. Lin², S.L. Teo², S. Tripathy², Armin Dadgar³, Alois Krost³; ¹Otto von Guericke University of Magdeburg, Germany ; ²IMRE, Singapore ; ³Otto von Guericke University of Magdeburg and AZZURRO Semiconductors AG, Germany.

A promising approach to achieve optical confinement in light emitting diodes (LED) is the fabrication of free standing micro-disk-emitters. Micro-disks LEDs of 300 &mum and 400 &mum diameter were characterized using Low temperature (4K) scanning u-photoluminescence (PL) as well as nano-cathodoluminescence (CL) spectroscopy. A standard 3 &mum thick GaN-LED structure grown by MOVPE on Si substrate and comprising a 5 x InGaN/GaN multiple quantum well (MQW) as active region was patterned into circular columns of 300 &mum and 400 &mum diameter using dry etching. Subsequently the Si substrate was partially removed using a selectively isotropically XeF2 enchant resulting in a distinctive undercut of the micro-disk LED. Local PL spectra at several spots on the sample show a dominant InGaN MQW emission peak at 2.66 eV strongly modulated by periodic Fairy-Perot-pattern as well as the GaN (D0,X) emission line at 3.46 eV. The InGaN-PL mode spacing of 83 meV corresponds to a "cavity" thickness of 3.2 &mum perfectly matching the total thickness of the nitride stack. On top of residual silicon stalk the MQW luminescence is red shifted directly visualizing the phase shift at the GaN/silicon as compared to the GaN/vacuum boundary. This is evidenced in a PL-spectrum-linescan across the whole diameter of a 300 &mum micro disk: well pronounced FPI patterns are visible showing an distinct abrupt spectral leap at the Si/vacuum boundary. Mapping of the GaN emission directly visualizes the 3D stress distribution in the warped micro-disk. A &mu-PL spectrum linescan across surface yields the lateral stress modulation resulting in a tensile stress of about 0.37 GPa at the free-standing part of the disk and 0.26 GPa on top of the Si stalk. Cross-sectional CL linescans from top to bottom of the disk exhibit the vertical stress relaxation on top of stalk as well as in the free standing disk region.

N4-S5.17
Microscopic Spatially Resolved Electrical and Optical Properties of MOCVD-Grown InGaN/GaN LEDs on Si(111) on Insulator. (#1075) Lars Reissmann¹, T. Fey¹, Armin Dadgar², Frank Bertram¹, S. Tripathy³, V.K.X. Lin³, S.L. Teo³, Juergen Christen¹, Alois Krost²; ¹Otto von Guericke University of Magdeburg, Germany ; ²Otto von Guericke University of Magdeburg and AZZURRO Semiconductors AG, Germany ; ³IMRE, Singapore.

InGaN/GaN-based light emitting diodes (LEDs) were grown by MOCVD on (111)-oriented 6 inch nanoscale silicon-on-insulator (SOI) substrates. Square-shaped mesa patterns are produced by multiple-mask photolithography, inductive coupled plasma etching, and contact metallization. Substrates for epitaxial growth are prepared by SIMOX process. The buried oxide layer has a thickness of 150 - 160 nm. In combination with the 45 nm thick overlaying final Si(111) an enhanced reflectivity may be attained for matched wavelengths due to Fabry-Perot effects. The LEDs were investigated by scanning micro-photoluminescence (u-PL) spectroscopy, and simultaneously laser beam induced current (LBIC) was detected. Additionally, on the same area of the LED scanning micro-electroluminescence (&mu-EL) was carried out under various forward currents. The combination of these characterization methods allows us to distinguish between different the injection luminescence determining mechanisms: &mu-PL shows a grainy distribution of InGaN luminescence intensity with a slight gradient from p-contact towards the n-contact region. In the wavelength image one can see larger blurry shapes of longer peak wavelengths (&Delta &lambda = 5 nm) compared to average. These spots have diameters of several tens of micrometers. As these shapes also appear at the same positions in &mu-EL, they can clearly be attributed to local thickness fluctuations and/or local fluctuations of refractive index - [In] fluctuation occurs on much smaller length scales. This is supported by local spectra taken from &mu-PL as well as &mu-EL mapping. These spectra exhibit different positions of Fabry-Perot interference modes (FP-modes). The LBIC signal gives additional access to electrical properties showing preferential current paths through the device similarly to &mu-EL mapping data but in the reverse direction. Thus, the influence of local surface properties of the LED can be distinguished from electrical properties of the bulk LED material. Local variations of surface properties such as dark spots (originating from contamination) or p-contact metallization geometry (bright border without contact metal) can be visualized. Non-uniform spatial distribution of injection current density is observed. Areas of lower current density show longer peak wavelengths. Higher injection current density near the n-contact region leads to blue shifted emission due to band filling.