Daniel Josell National Institute of Standards and Technology
Michael Brett University of Alberta
Christian Witt AMD
c/o IBM T. J. Watson Research Center
Mikko Ritala University of Helsinki
D1/AA1: Joint Session: Fabricating and Filling Complex 3-d Structures
Tuesday PM, April 10, 2007
Room 3003 (Moscone West)
9:30 AM - **D1.1/AA1.1
Electrodeposition Through Colloidal Templates: Fabrication of Nanostructures, Properties and Applications.
Philip Bartlett 1 Show Abstract
1 Chemistry, Southampton University, Southampton, Hampshire, United Kingdom
Templated electrochemical deposition through close packed monolayers of unform polystyrene colloidal particles assembled on electrode surfaces produces structured thin films. Removal of the template by dissolution leaves a supported thin metallic film containing an array of interconnected spherical segement voids. Because electrodeposition is a volume filling technique which does not require a heat treatment step the diameters and organisation of these voids replicates the diameter and packing of the colloidal particles used to form the template. In addition the thickness of the film is controlled by the charge passed to deposit the film. It is thus possible to simply and predictably control the geometry of the structured films produced. These structured metal films have interesting magnetic , superconducting  and optical properties  that are determined by their precise geometry.Using templates with diameters between 450 and 1200 nm we are able to produce metallic surfaces that show significant surface enhancement in surface enhanced Raman spectroscopy (SERS) if the correct sphere diameter and film thickness are used. We have investigated the origins of this surface enhancement by varying the film thickness, template sphere diameter and by looking at the angular dependence. We find that the intensity of the SER spectra varies with all of these factors indicating that the precise geometry of the structured surface and the excitation of surface plasmons is important in producing the surface enhancement. Studies of the SERS intensity at the structured surfaces show that the signal is linear with laser light intensity. We also find that the enhancement varies with the laser wavelength. These results are consistent with electromagnetic enhancement of the SERS signal caused by the excitation of confined plasmons  at the structured metal surface. This tentative explanation is consistent with our detailed studies of the reflection spectra of the structured metal films as a function of pore diameter, film thickness and angle of incidence . In contrast to the electrochemically roughened surfaces the intensity of the SERS spectra on the structured surface is reproducible from place to place across the surface and from sample to sample. This is a significant potential advantage. The structured surfaces are also robust and stable under laboratory conditions and ideally suited as electrodes for electrochemical SERS experiments since they do not have a high surface area. References:  P. N. Bartlett, M. A. Ghanem, I. S. El Hallag, P. de Groot, A. Zhukov, J. Mater. Chem., 13, 2003, 2596. A. A. Zhukov, E. T. Filby, M. A. Ghanem, P. N. Bartlett and P. A. J. de Groot, Physica C, 404, 2004, 455. P. N. Bartlett, J. J. Baumberg, S. Coyle and M. Abdelsalem, Faraday Discuss., 125, 2004, 117. S. Coyle, M. C. Netti, J. J. Baumberg, M. A. Ghanem, P. R. Birkin, P. N. Bartlett and D. M. Whitaker, Phys. Rev. Lett., 87, 2001, 176801-1.
10:00 AM - D1.2/AA1.2
Variable Filling Fraction Inverse Opal Metallic Photonic Crystals
Xindi Yu 1 , Yun-Ju Lee 1 , Robert Furstenberg 2 , Jeffrey White 2 , Paul Braun 1 Show Abstract
1 Material Science and Engineering, University of Illinois, Urbana, Illinois, United States, 2 Physics, University of Illinois at Urbana Champaign, Urbana, Illinois, United States
Metallic photonic crystals, metal based structures with periodicities on the scale of the wavelength of light, have attracted considerable attention due to the potential for new properties, including the possibility of a complete photonic band gap with reduced structural constraints compared to purely dielectric photonic crystals, unique optical absorption and thermally stimulated emission behavior, and interesting plasmonic physics. Photonic applications may include high efficiency light sources, chemical detection, and photovoltaic energy conversion. Other applications for three-dimensionally porous metals, so called “metal foams”, include acoustic damping, high strength to weight structures, catalytic materials, and battery electrodes. The photonic properties of metallic inverse opal structures have been of significant interest because of the simplicity of fabrication and potential for large area structures. However, in practice, experiments on metal inverse opals have been inconclusive, presumably because of structural inhomogeneities due to synthetic limitations. In this work, we demonstrate an electrochemical approach for fabricating high quality metal inverse opals with complete control over sample thickness, surface topography and for the first time, the structural openness (metal filling fraction). Optical measurements conclusively demonstrate that metal inverse opals modulate the absorption and thermal emission of the metal and that these effects only become three-dimensional (3D) in nature at high degrees of structural openness.
10:15 AM - **D1.3/AA1.3
Surface Engineering of Aerogels via Atomic Layer Deposition
Juergen Biener 1 Show Abstract
1 Nanoscale Synthesis and Characterization Laboratory, Lawrence Livermore National Laboratory, Livermore, California, United States
Nanoporous materials with tailored surface functionality hold technological promise for applications such as sensors, energy storage, and catalysis. Such materials can be developed using aerogels as a flexible and robust material platform. While macro-cellular open-cell foams can be easily coated via chemical or physical vapor deposition, shadowing effects and diffusion limitations become dominant at the submicron length scale. I will discuss a general approach to prepare such metal/aerogel nanocomposites via template directed atomic layer deposition (ALD). The approach was tested for a wide range of aerogel templates and ALD processes including the deposition of W, Ru, Pt, Cu, CuN and ZnO. In general, the process offers excellent control over composition and morphology. Furthermore, the ability to control the growth morphology of the deposited material opens the door to further fine-tune material properties by exploiting the size-effect frequently observed for nanoparticles. I will also discuss the limitations of the ALD approach, and suggest ways to overcome these.
11:15 AM - **D1.4/AA1.4
New Routes to Three-Dimensional Photonic Band-Gap Materials
Martin Hermatschweiler 1 , Sean Wong 2 , Alexandra Ledermann 2 , Geoffrey Ozin 4 , Martin Wegener 1 2 3 , Georg von Freymann 1 2 3 Show Abstract
1 Institut für Angewandte Physik, Universität Karlsruhe (TH), Karlsruhe Germany, 2 Institut für Nanotechnologie, Forschungszentrum Karlsruhe, Karlsruhe Germany, 4 Chemistry Department, University of Toronto, Toronto, Ontario, Canada, 3 DFG-CFN, Universität Karlsruhe (TH), Karlsruhe Germany
The past decade has witnessed intensive research efforts related to the design and fabrication of Photonic Crystals (PCs) [1,2]. These periodically structured dielectric materials can represent the optical analogue of semiconductor crystals and provide a novel platform for the realization of integrated photonics. While the layer-by-layer or “woodpile” structure  is amenable to microfabrication techniques and has already been realized at infrared frequencies through combinations of advanced planar semiconductor microstructuring techniques for individual layers with sophisticated alignment and stacking procedures to combine different layers into 3D PCs , more demanding structures like, e.g., photonic quasicrystals or chiral photonic crystals, cannot directly be realized with conventional techniques. Here we present two alternative approaches, which allow the realization of almost arbitrary three-dimensional structures: (i) In a first step, a three-dimensional polymer template is created via direct-laser-writing  (DLW). As the optical properties, especially the index of refraction, of the polymer (SU-8) are not sufficient for the opening of a complete photonic bandgap, we use a combination of atomic-layer-deposition (ALD) of silica and chemical vapor deposition (CVD) of silicon to replicate  or invert these intricate 3D templates with silicon. (ii) To overcome the need for additional ALD and CVD techniques, we directly write 3D structures into a high-index of refraction all-inorganic photoresist , namely into As2S3 chalcogenide glass. With a specially designed highly selective wet chemical etch, the unexposed areas are removed. Besides its high index of refraction, As2S3 is advantageous for 3D nanofabrication as it does not swell or shrink during etching, as polymer materials do. E. Yablonovitch, Phys. Rev. Lett. 58, 2059-2062 (1987).  S. John, Phys. Rev. Lett. 58, 2486-2489 (1987). K.-M. Ho et al., Solid State Comm. 89, 413-416 (1994). S. Y. Lin et al., Nature 394, 251-253 (1998).  M. Deubel et al., Nature Mater. 3, 444 (2004). N. Tétreault et al., Adv. Mater. 18, 457 (2006). S. Wong et al., Adv. Mater. 18, 265 (2006).
11:45 AM - **D1.5/AA1.5
Nano-Fabrication of 3D Optoelectronic Devices by Atomic Layer Deposition
Christopher Summers 1 , Elton Graugnard 1 , Davy Gaillot 1 , John Blair 1 , Olivia Roche 2 , David Sharp 2 , Robert Denning 3 , Andrew Turberfield 2 Show Abstract
1 Materials Science and Engineering, Georgia Institute of Technology, Atlanta, Georgia, United States, 2 Department of Physics, University of Oxford, Oxford United Kingdom, 3 Inorganic Chemistry Laboratory, University of Oxford, Oxford United Kingdom
The formation of 3D periodic dielectric contrast allows great control over the propagation of light. The two main challenges in fabricating patterned dielectric materials are pattern formation and dielectric control. Atomic layer deposition (ALD) has proven to be a powerful tool for fabricating new dielectric structures from patterned templates (produced by a variety of techniques) by enabling their facile inversion and replication into a wide range of dielectric materials. [1,2] Here, we present ALD of high transparency, high index, luminescent, electro-optic and nonlinear materials into self-assembled opal and lithographically derived templates, where each presents unique challenges and benefits as template structures. Investigations are presented on the fabrication of inverse opal based structures by multi-layer ALD of TiO2, Al2O3, and ZnS. Most recently new protocols have been developed for the ALD growth of GaP on metal oxide templates. It is demonstrated that novel derivatives of the opal structure can be obtained by the use of a sacrificial buffer layer, which is subsequently removed with the original template by etching and followed by regrowth of the target dielectric material. Consequently, structures can be inverted, precisely replicated, and formed from composite or multi-layered materials that allow a high degree of functionality: for example, luminescence modification and photonic band tuning. Additionally, this process enables polymer structures formed at low temperature to be inverted by low temperature ALD into a high temperature compatible material that then serves as a second high temperature template (double templating). Specific examples are given of the high degree of static and dynamic tunability (by liquid crystal infiltration) that can be obtained in non-close-packed opals, and recent work presented on the inversion and replication of holographically defined SU-8 templates and biological scaffolds. This work demonstrates a pathway to the fabrication of new lattices with complex geometries and material combinations that are needed to enable full control over light propagation in dielectric material devices. J. S. King, E. Graugnard, O. M. Roche, D. N. Sharp, J. Scrimgeour, R. G. Denning, A. J. Turberfield and C. J. Summers, Adv. Mater., 18, 1561 (2006). E. Graugnard, J. S. King, D. P. Gaillot and C. J. Summers, Adv. Func. Mater., 16, 1187 (2006).
12:15 PM - D1.6/AA1.6
Periodic Nanostructures Templated from Two-Dimensional and Three-Dimensional Colloidal Crystals
Peng Jiang 1 Show Abstract
1 Chemical Engineering, University of Florida, Gainesville, Florida, United States
Self-assembled colloidal crystals are ideal templates for creating three-dimensional (3D) highly ordered macroporous materials and photonic crystals. In this approach, the voids between colloidal spheres are infiltrated with another material and subsequent removal of the template by either wet etching or thermal decomposition leads to the formation of 3D ordered air cavities inside the structure-filling materials. Here we report two nonlithographic approaches to the fabrication of large-area, periodic nanostructured materials using both 2D and 3D colloidal crystals as templates. In the first approach, 3D ordered colloidal crystals, made by convective self-assembly or spin-coating, are used as templates to create 2D surface gratings, provided the subconformal deposition of materials by physical vapor deposition only occurs on the periodic surfaces of the templates. The technique allows the production of surface gratings from a large variety of functional materials, such as metals, semiconductors, and dielectrics. In the second approach, two-dimensional non-close-packed colloidal crystals fabricated by a simple spin-coating process are used as structural templates to make metallic nanohole arrays. Wafer-scale samples (up to 8-inch) can be easily made using standard physical vapor deposition. Complex micropatterns can be created by standard microfabrication for potential device applications. The templated periodic nanostructured materials have important technological applications in subwavelength optics, plasmonic sensors, and efficient organic light-emitting diodes (OLEDs).
12:30 PM - D1.7/AA1.7
Effect of SiC Whisker Growth on Cordierite Honeycomb by CVI Process.
Hwan Sup Lee 1 , Ik Whan Kim 1 , Doo Jin Choi 1 , Hia Doo Kim 2 Show Abstract
1 Department of Ceramic Engineering, Yonsei University , Seoul Korea (the Republic of), 2 Department of Materials Engineering, Korea Institute of Machinery and Materials, Seoul Korea (the Republic of)
Recently, there has been a increasing concern for the air pollutant caused by diesel engine such as particulates and nitrogen oxides. As these pollutant are mainly caused by diesel automobiles. Legal regulations are becoming rigorously strengthened. Hence, the diesel particulate filter (DPF) made of ceramic material is developed, to eliminate particulates generated by the diesel engines. For the application of DPF, the silicon carbide DPF with high strength are mainly used, and in developed countries, it is a standardized regulation to use these silicon carbide DPFs in the cutout units of diesel cars. However, the silicon carbide DPFs currently available in the market have 10μm mean pore size, which has also efficiency problem to capture nano-particulates less than 50nm size, and it is insufficient capturing filter to achieve the ultimate goal - i.e. atmosphere environment protection by capturing the particulates. This study was performed in order to improve conventional DPF efficiency by filtering nano-particle materials. SiC whiskers were grown on porous cordierite honeycomb in order to enhance the filtering efficiency, performance, and durability by controlling pore morphology. This experiment was performed by chemical vapor infiltration(CVI) process in order to grow the whiskers at inner pores without closing the pores. Metyltrichlorosillan(MTS) was used as source gas, and the input gas ratio α[H2/MTS] were 100 and 300, respectively. the deposition temperature was varied from 1000°C to 1300°C, and the deposition time was varied from 10min to 120min. After the deposition process, scanning electron microscopy(SEM) and universal testing machine were used to explain the micro-structure and compressive strength, respectively. The sizes of pores were measured with mercury porosimeter, and the gas permeability was analysed by the nitrogen gas injection method under 1 atmospheric pressure at 28°C, and the particulates trap test was conducted. The formation of "networking structure" in the inside of porous cordierite honeycomb was confirmed. The compressive strength of whiskered porous cordierite honeycomb increased from 24MPa to 60MPa(250%) after 1hour deposition at 1200°C. And the permeability of whiskered porous cordierite honeycomb was higher than that of cordierite honeycomb with similar pore size. This study showed the advantage of networking structure by whiskers in order to filter the nano-particulate materials with improved strength, reduced pore size and minimized permeability drop.
D2: Feature Filling in Supercritical Fluids and Electrolytes
Tuesday PM, April 10, 2007
Room 3003 (Moscone West)
2:30 PM - **D2.1
Conformal Deposition of Metal, Metal Oxide and Mixed Metal Oxide Films within Nanostructured Device Features from Supercritical Fluids.
Adam O'Neil 1 , Christos Karanikas 2 , James Watkins 1 Show Abstract
1 Polymer Science and Engineering, University of Massachusetts, Amherst, Massachusetts, United States, 2 Chemical Engineering, University of Massachusetts, Amherst, Massachusetts, United States
Reactive deposition from supercritical CO2 yields device quality films with exceptional step coverage in high aspect ratio, nano-scale features in a single step. While the deposition chemistry is similar in some respects to CVD, the use of CO2 solutions as the reaction media enables precursor transport in solution at volumetric concentrations many orders of magnitude higher than those employed in CVD while maintaining the benefits of gas-like transport (no surface tension, high diffusivity). The high concentration of available precursor results in surface reaction rate limited, zero order deposition kinetics (with respect to precursor), very good deposition rate (e.g. 20 – 50 nm / minute for Cu) and exceptional step coverage. The use of cold-wall reactors equipped with heated platens yields substrate selective depositions and high precursor conversions (90 + %). In this talk the conformal deposition of metal, metal oxide and mixed metal oxide films within 2-D and 3-D nano-structured device features will be discussed. The basis for conformal coverage is rationalized through detailed studies of the deposition kinetics. Recent extensions of the technique for conformal deposition of single and mixed metal oxide films including hafnia and other high k materials by a clean precursor hydrolysis mechanism at temperatures below 400 0C will be highlighted. The ability to deposit metal and metal oxide films in topographically challenging structures offers numerous opportunities for next generation devices. A number of examples will be discussed. For example, conformal metal electrode deposition combined with the deposition of conformal high k oxides offers direct routes to high aspect ratio capacitors on chip and for non-volatile memory applications.
3:00 PM - D2.2
Detailed Kinetics and Conformal Coverage in High Aspect Ratio Features for the Deposition of Ruthenium from Supercritical Carbon Dioxide.
Christos Karanikas 1 , James Watkins 2 Show Abstract
1 Chemical Engineering, University of Massachusetts Amherst, Amherst, Massachusetts, United States, 2 Polymer Science and Engineering, University of Massachusetts Amherst, Amherst, Massachusetts, United States
3:15 PM - D2.3
Novel “Topography-sensitive” Bottom-up mode for Filling Nano-features Using Superciritical Fluids as a Deposition Medium.
Eiichi Kondoh 1 , Michiru Hirose 1 Show Abstract
1 , Univ. Yamanashi, Kofu Japan
A supercritical fluid is any substance at a temperature and pressure above its thermodynamic critical point. It has the unique ability to diffuse through solids like a gas, and dissolves materials like a liquid[Wikipedia]. We have developed technology of depositing metallic thin films (Cu, Pt, Pd, etc.) using supercritical carbon dioxide fluids as a deposition medium, proving excellent gap-filling capability, low temperature deposition possibility, and a cleaning role of the medium fluid. In this paper, a novel success in using supercritical fluids as a deposition medium, so-called “topography-sensitive” deposition, is proposed and demonstrated. Supercritical fluids are dense media, and a deposition precursor dissolving wherein has a partial pressure as high as, for instance, 0.1 MPa. This value is much higher than the precursor itself’s vapor pressure, and thus such a high pressure is hardly achieved by other common vapor or wet processes. Under such a condition, the condensation of precursor can easily takes place especially in a concave feature such as holes or trenches. Therefore, by combining this phenomenon with proper reaction chemistry, it can be realized to deposit the target material in the concave structures preferentially. The concept is completely different from usual “selective deposition” in CVD or electrochemical deposition, where the selectivity is originated from the difference of reactivity or nucleation-promoting ability of the substrate surface. On the contrary, the origin of our topography-sensitive mode is simply the topography of the substrate surface. In principle, the narrower the feature in size, the better condensation takes place. That is, deposition prefers in narrower features to occur. This is a possible reason why very narrow features can be filled without seam whereas conformal deposition proceeds the larger features. This unique property can provide possibilities of, such as, sub deca-nano interconnect formation. Cu and Ru were selectively deposited in trenches in a “topography-sensitive” manner. Note that the underlying surface is a refractory barrier metal conformally deposited using an atomic-layer-deposition technique.
3:30 PM - D2.4
Quantitative Evaluation of Gap-filing and Conformal Deposition in High Aspect Ratio Features by Supercritical Fluid Deposition of Copper.
Takeshi Momose 1 , Masakazu Sugiyama 2 , Yukihiro Shimogaki 1 Show Abstract
1 Department of Materials Engineering, The University of Tokyo, Tokyo Japan, 2 Institute of Engineering Innovation, The University of Tokyo, Tokyo Japan
Supercritical fluid deposition (SCFD), which is a reduction of metal organic compounds with H2 in supercritical carbon dioxide (scCO2), is reported to be a promising technology for superior gap filling and structure-independent conformal deposition. A possible reason of such property is that large precursor concentration in scCO2 realizes Langmuir-Hinshellwood-type nonlinear surface reaction kinetics in which the deposition rate is independent of the precursor concentration. In addition, high precursor concentration accomplishes high nucleation density suitable for the deposition of very thin conformal films. For example, we realized 10-nm thick ultra thin continuous Cu film fabrication on Ru substrate using SCFD. To date, however, there has been no quantitative evaluation on the potential of the conformal deposition by SCFD. In chemical vapor deposition (CVD) and atomic layer deposition (ALD), on the other hand, systematic evaluation of gap-filling property has been developed. The step coverage, which is the ratio of the film thickness between the top and the bottom of high aspect ratio structures, has clear dependence on the aspect ratio, which provides a lot of quantitative information on the surface kinetics. We can even estimate the intrinsic limit of the gap-filling or conformal deposition based on such information.Because the SCFD has high potential of conformal deposition, in order to observe the deviation of the step coverage from unity, we have to observe the deposition in extremely narrow structures with high aspect ratio, which has been a technological challenge. We employed a novel method, angled polishing of patterned substrates, in order to evaluate the step coverage of the Cu films obtained by SCFD for via holes with various diameters and hence various aspect ratios. This method enables us to observe the cross-sectional plain-view SEM images of via holes at continuous depth from the top to the bottom.It was confirmed that our typical deposition condition enabled seamless and void-free gap-filling into the vias with the aspect ratio of 20 (50 nm in diameter). In such perfect gap-filling, the remaining hole continuously shrinks and the aspect ratio increases accordingly. Therefore, the SCFD process we examined realizes unity step coverage for infinitely large aspect ratio. It is also possible that some mechanism other than surface reaction kinetics enabled such super-filling. Time dependent analysis of the step coverage, and the observation of the effect of the deposition condition, such as temperature, the concentrations of a precursor and H2, provide key information to elucidate the mechanism of excellent gap-filling and conformal deposition, and finally to clarify the technological limit of the filling property of this SCFD process.
4:15 PM - **D2.5
Chemical Processing for On-Chip Interconnects.
Valery Dubin 1 Show Abstract
1 , eMAT Technology, Moses Lake, Washington, United States
4:45 PM - **D2.6
Electroless Deposition for ULSI Cu Interconnects Application
Tetsuya Osaka 1 , Madoka Hasegawa 1 , Masahiro Yoshino 1 Show Abstract
1 Dept. of Applied Chemistry, Waseda University, Tokyo Japan
Copper interconnects have been fabricated by so-called Damascene process, which is trench-filling process by copper electrodeposition. Although, this process achieved void-free filling of high aspect-ratio submicrometer trenches, some critical problems, such as subconformal deposition of sputtered seed and barrier layers and uneven distribution of current on a wafer, still remain to be addressed. Electroless deposition is expected to be useful for fabricating Cu interconnects because it is capable of depositing thin and uniform metal layer on high aspect-ratio nano-structures fabricated even on large substrates. This presentation reviews our recent results on two electroless deposition processes both for ultra large-scale integration (ULSI) application. One process is a copper electroless deposition process for filling of damascene trench-structures. Another one is a Ni alloy electroless deposition process for fabricating barrier layer. Electroless deposition was applied to copper filling process. Electroless deposition is believed to be effective for filling of further miniaturized trench structures because it dose not require thin sputtered seed layer although an appropriate catalyzation is required. For this process, copper should be filled in trenches without producing voids. To achieve void-free filling, or “superfilling”, electroless Cu deposition for the filling of trenches was investigated using polyethylene glycol (PEG) as an inhibiting additive. Superfilling of 100 nm-wide and 300 nm-deep trenches was achieved by the addition of appropriate concentrations of PEG. The significant bottom-up growth in trenches was observed with this bath. Electroless deposition was also applied to the formation of diffusion barrier layer. For this process, Ni alloy was employed as a barrier layer material. We applied Pd-catalyzed self-assembled monolayer (SAM) to the adhesion/catalysis layer. This technique was the key for achieving uniform thin layers with sufficient adhesion to both silicon dioxide and low dielectric constant substrates. We demonstrate that this electroless deposition process is effective for depositing conformal films on the high aspect-ratio trenches, compared with conventional process such as spattering. Finally, electroless NiB film with 20 nm thickness was successfully formed on trench-patterned substrate with excellent step coverage. This barrier layer also exhibited sufficient barrier properties against copper diffusion into the substrate and acceptable thermal stability. We have achieved superfilling of electrolessly deposited Cu, and have successfully developed the process for forming electroless Ni alloy barrier layer using Pd-activated SAM. The possibility of fabricating interconnecting layers by all electroless process is of great interest. The attempt to combine these two processes will be also discussed.
5:15 PM - **D2.7
Advanced Cu Electrodeposition Technology for Next Generation Damascene Metallization.
Jae Jeong Kim 1 , Sung Ki Cho 1 Show Abstract
1 School of chemical and biological engineering, Seoul National University, Seoul Korea (the Republic of)
Cu electrodeposition is regarded as the appropriate Cu deposition method in damascene metallization. In sub-micron technology generation, Cu electrodeposition has only been expected to superfill the trenches by using organic additives. However, as the line width of trench decreases sharply below 100 nm, other issues arise. First of all, this decrease makes the loss of a fraction of conductor in line due to barrier/seed layer. Direct electrodeposition on barrier (without seed) is a promising solution for this problem and the various barrier layers such as Ta, Ru, Os, and Ir have been tested for it and many researched have been performed actively. Besides the increase in conductor fraction, the replacement of Cu by other metals is suggested in order to improve the electrical, chemical and physical properties of conductor. Ag and Au electrodeposition is attempted in this sense and some impressive superfilling of Ag and Au has been reported. Next issue is the formation of topographic variations on wafer surface. Topographic variation is formed due to the function of organic additives, and it happens obviously at small pattern. It can be translated into the variations of metal thickness after CMP. Therefore, leveling, which is the reduction of the variations during electrodeposition, is issued recently. A few leveling methods have been suggested. Last issue is about the reliability of Cu related to electromigration. Cu surface is vulnerable to electromigration and it becomes severe with decrease in line width and following increase in maximum current density of circuit. Therefore, Cu surface should be covered with thin film, called capping layer, to prevent the Cu diffusion through the surface. Co film doped with W and P is known as the effective capping material and the selective capping of Co(W)P has been reported. In this presentation, we cover four issues, that is, direct electrodeposition, superfilling, leveling, and capping layer. In direct electrodeposition, we present the Cu deposition on barrier layer of TiN and Ru. At both, the surface treatment is essential for the formation of continuous film because of the low nuclei density of Cu on the high-resistive barrier layer. In addition, Cu superfilling is studied with various additives and each filling mechanism is suggested. And Ag superfilling is introduced with organic additives of BTA and thiourea which are the unique combination for superfilling. Also, superfilling can be achieved without additives and some novel methods using self-assembled monolayer or anodic current are presented herein. In leveling, we exhibit the two kinds of leveling methods. One is using organic leveler of BTA to flatten the Cu surface. The other is based on the control of the adsorption of organic additive by change of applied potential and without leveler. Co(W)P capping layer to improve the Cu reliability is also studied. Co(W)P layer is deposited electrolessly with the optimized solution composition.
5:45 PM - D2.8
Bottom-up Feature Filling with Mobile Adsorbates: Contrasting Cu, Ag and Au Superfilling of Sub-micrometer Trenches and Vias.
Daniel Josell 1 , Thomas Moffat 1 , Daniel Wheeler 1 Show Abstract
1 Metallurgy Division, NIST, Gaithersburg, Maryland, United States
Highly nonconformal bottom-up filling of trenches and vias known as “superfilling” dominates state-of-the-art interconnect fabrication, being a key aspect of the Cu electrodeposition processes used for on-chip interconnect fabrication. The bottom-up filling has been quantitatively described by the Curvature Enhanced Accelerator Coverage (CEAC) mechanism that explains how local changes of surface area affect the coverage of deposition rate accelerating adsorbates derived from additives in superfilling electrolytes. CEAC-based models have quantitatively explained industrially relevant Cu superfilling as well as Ag and Au superfilling, all during electrodeposition, as well as Cu superfilling during catalyzed chemical vapor deposition. It is implicit in existing CEAC models that, while the adsorbates are surfactant in nature and thus can “float” on the surface during the course of metal deposition, they do not diffuse laterally along the deposit surface; such mobility would lead to smoothing or, in extreme limits, elimination of the gradients of adsorbate coverage that arise from the CEAC mechanism and underlie superfilling processes. This work incorporates surface mobility into the CEAC framework and then shows how differences between surface contours observed during trench and via filling through disclosed processes for Cu, Ag and Au superfilling can be explained utilizing the new construct as arising from differences in the impact of adsorbate mobility on CEAC induced superfilling for the different systems.
Daniel Josell National Institute of Standards and Technology
Michael Brett University of Alberta
Christian Witt AMD
c/o IBM T. J. Watson Research Center
Mikko Ritala University of Helsinki
D3: ALD and CVD for Trenches, Vias and Pillars
Wednesday AM, April 11, 2007
Room 3003 (Moscone West)
9:30 AM - **D3.1
Issues with Narrow Trench Metallization.
Sywert Brongersma 1 , L. Carbonell 1 , Z. Tokei 1 Show Abstract
1 , IMEC, Leuven Belgium
The continued scaling of interconnect structures poses technological challenges is several fields of expertise. Clearly making increasingly small trenches in porous dielectrics is a challenging task by itself and several routes to making relevant sub 50 nm high aspect ratio test structures for filling will be shortly highlighted.Metallizing such structures consists of three main steps, each with their own specific issues that will be highlighted consecutively. Firstly the barrier material is typically sputtered into the feature and this approach continues to demonstrate extendibility to much more aggressive nodes than was foreseen through extensive step-coverage optimization. However, combinations of sputtering with ALD, combining more traditional Ta based barriers with e.g. Ru, are being explored as well.For the seedlayer sputtering is still the mainstream approach as well, but direct plating on more advanced barrier stacks is being explored in order to eliminate this step. Finally, plating is still being extended to increasingly challenging structures. Here the additives in the plating solution need to be optimized to maximize curvature enhanced superfill behavior while minimizing overshoot and structure dependent variations. Also, a balance between impurity induced resistivity increases and reliability enhancement is being explored.In conclusion, a wide variety of issues in the several metallization steps are being addressed for future technology nodes.
10:00 AM - **D3.2
Deposition of Conformal Thin Films by Atomic Layer Deposition.
Markku Leskela 1 , Marianna Kemell 1 , Markus Lautala 1 , Tero Pilvi 1 , Viljami Pore 1 , Eero Santala 1 , Mikko Ritala 1 Show Abstract
1 Department of Chemistry, University of Helsinki, Helsinki Finland
Because of the self-limiting growth principle Atomic Layer Deposition (ALD) suits very well for depositing thin films on nonplanar surfaces. Neither the size of the substrates nor the size of the structural features has any clear threshold value in ALD. ALD suits very well for making and modifying of nanomaterials. In this presentation the use of ALD for depositing conformal films on nonplanar surfaces is exemplified by three ways: deposition of thin films on nanostructured high surface area substrates, deposition films on inside walls of through porous substrates, using nanofibres and nanospheres as templates in preparation of nanotubes and hollow nanospheres. The materials deposited by ALD and presented here are mostly oxide (Al2O3, TiO2) and metal (Ru, Ir, Pt) films. High surface area substrates are of interest as capacitors as known in trench structure DRAM devices. Through-substrate holes coated with a metal film represent one interesting application with great application potential also in microelectronics. Nanosized membranes are important in separations and by modifying the wall surface the application area can be extended. The possible use of nanotubes and nanospheres is still largely to be discovered but they may be applied for example in sensors, drug release, and fluid transportation.
10:30 AM - **D3.3
Control of Step Coverage in Atomic Layer Deposition (ALD): Theory and Applications
Roy Gordon 1 Show Abstract
1 Chemistry and Chemical Biology, Harvard University, Cambridge, Massachusetts, United States
ALD is usually presented as a technique that produces completely conformal coatings. In fact, ALD process conditions can be set to produce any desired degree of conformality. The conformality of ALD coatings varies according to a simple formula relating the aspect ratio approximately to the square root of the exposure to the precursor vapor. Exposure is defined as the partial pressure of a precursor vapor integrated over the time that the vapor remains at the open end of a hole or trench. Examples of the same chemistry are shown to produce either conformal coating inside holes with high aspect ratio using high exposure, or coating just at the open ends of holes under low exposure. Deviations from complete conformality can also result from non-ideal surface chemistry, such as thermal decomposition, competitive adsorption of byproducts, or etching by reactants or byproducts. Reactions showing each of these non-ideal behaviors will be reviewed. Highly conformal ALD coatings are produced commercially in DRAM memory elements, read/write heads in magnetic disk memories, in optical phase plates and polarizers, and on irregular phosphor particles. Potential future uses of conformal ALD coatings include photonic crystals, MEMS devices, and coatings of heterogeneous catalysts. Non-conformal ALD coatings can be used for sealing pores in porous dielectrics or other porous materials.
11:30 AM - D3.4
Novel ALD WCx Process for High Aspect Ratio Structure.
Wei-Min Li 1 , Eva Tois 1 , Kurt Verheyden 2 , Jeroen van Hapert 2 Show Abstract
1 , ASM Microchemistry Ltd, Helsinki Finland, 2 , ASM Belgium N.V., Leuven Belgium
Tungsten carbide is known to have low resistivity, high thermal and chemical stability, superior hardness, low friction and catalyst effects. These properties are desired for many demanding applications with substrates having complex geometry, such as diffusion barrier for interconnects in damascene or 3-D integration, metal electrode for memory and logic applications, wear resistant MEM devices as well as catalyst. Literature reported WC depositions often involve high energy sources such as plasma or very high temperature. Low temperature CVD using metal-organic precursors were also reported but impurities were fairly high. Report on conformal WCx deposition in high aspect ratio substrates has not been found.We have discovered a novel process of WCx by thermal ALD. The thin film can be deposited at temperatures as low as 200 to 300 οC. The growth rate at 250 οC process temperature was 0.8-0.9 Å/cycle. XPS showed the film was of high purity with atomic ratio of W:C=55:45. Impurities were not observed within their detection limit except Si (1 at-%) likely due to the signal contribution from Si substrate used. Typically, the as-deposited WCx thin film on 20 nm thermal SiO2 covered 200 mm Si wafer has a sheet resistance non-uniformity of less than 3% (1 σ), as measured by four-point-probe (49 points with 3 mm wafer edge exclusion). By measuring the film thickness via XRR and spectroscopic ellipsometer, the converted bulk resistivity of as deposited thin films was approximately 400 μΩcm. XRD analysis showed a broad band with peak overlaps that of WC1-x, indicating a nanocrystalline and amorphous texture. No phase change was observed by XRD upon rapid thermal annealing in N2 or He up to 850 οC, however, the resistivity of the annealed samples decreased from 430 to 270 μΩcm. Upon further annealing at 1000 οC, phase change from WC1-x to WC was observed and the resistivity was increased back to 430 μΩcm. WCx film deposited on Al2O3 and HfO2 thin films behaves similarly upon annealing. In order to utilize the advantage of conformal coating by ALD, we have deposited WCx in high aspect ratio (60:1) trench wafers with line width of 110 nm and depth of 6 μm. SEM cross section revealed that step-coverage of greater than 80 % can be achieved.In conclusion, a novel low temperature thermal ALD WCx process has been developed with potential applications such as Cu diffusion barrier, metal electrode, and metal gate materials. In particular, the ability of depositing conformal WCx thin film in high aspect ratio structures makes it attractive for applications on nonplanar substrates where a conformal, thermally and chemically stable, wear resistant, or low resistivity metallic thin film is required.
11:45 AM - D3.5
CVD Controlled by Surface Chemistry: Superconformal Coverage and Filling of Deep Trenches and Vias
John Abelson 1 , Yu Yang 1 , Sreenivas Jayaraman 1 , Do-Young Kim 2 , Navneet Kumar 1 , Sophia Lazarz 1 , Wontae Noh 2 , Gregory Girolami 2 Show Abstract
1 Materials Science and Engineering, U. Illinois at Urbana-Champaign, Urbana, Illinois, United States, 2 Department of Chemistry, U. Illinois at Urbana-Champaign, Urbana, Illinois, United States
In the chemical vapor deposition of thin films, excellent conformal coverage on high aspect ratio trenches or vias occurs when the surface reaction probability is extremely low, e.g. < 10-4 for depth:width ratios of > 20:1. We show that this can be achieved at low substrate temperatures (200-300°C) using an appropriately designed precursor molecule, and can be enhanced using a second species that acts as a growth suppressor. The control mechanism is site blocking, wherein the surface becomes largely passivated by adsorbed precursor molecules, dissociated ligands, or the suppressor species. Utilizing this approach, we have achieved perfect conformal coverage, complete filling, and superconformal (bottom up) growth of thin films in trenches and vias with aspect ratios > 20:1. The method is a steady-state CVD process with no time variation of the injected fluxes, substrate temperature, or other variable; the growth rates are significantly faster than achievable by ALD. We report the film coverage profile, growth rate, composition and properties as a function of precursor pressure and substrate temperature for the refractory conductor HfB2, the refractory dielectric MgO, and elemental Ru. The HfB2 films are grown using the borohydride precursor Hf(BH4)4; the MgO and Ru films employ newly-synthesized precursors that we will discuss at the symposium. We analyze the reaction kinetics using desorption mass spectrometry, spectroscopic ellipsometry, and the thickness profiles on trench substrates with 100:1 aspect ratios. Extremely conformal coverage is achieved in the limit of low temperatures and high precursor pressures, which jointly maximize the site blocking effect. Superconformal growth, in which the film grows at a faster rate deep within the trench or via than near the opening, is obtained using a suppressor species. The suppressor must have a higher sticking probability than that of the precursor molecule and must prevent the adsorption or growth reaction of the precursor. Under these conditions, film growth rate is suppressed only on surfaces that are the most directly exposed to the incident flux, i.e. near to the opening. We show that atomic hydrogen is a useful suppressor for CrB2 growth from the precursor Cr(B3H8)2; that atomic nitrogen is suitable for HfB2 growth; and that molecular suppressors can also be utilized. Our results demonstrate that the control of surface chemistry provides a practical means to achieve conformal coatings on very high aspect ratio features using steady-state CVD. We suggest that the use of suppressors can be generalized to achieve conformal and super-conformal film growth for a wide variety of precursors.
12:00 PM - D3.6
Analysis and Simulation of Conformal and Super-Conformal CVD on High Aspect Ratio Features.
Yu Yang 1 , Gregory Girolami 2 , John Abelson 1 Show Abstract
1 Materials Science and Engineering, U. Illinois at Urbana-Champaign, Urbana, Illinois, United States, 2 Dept. of Chemistry, U. Illinois at Urbana-Champaign, Urbana, Illinois, United States
The need to coat or fill recessed features such as trenches or vias is frequently encountered in micro- and nano-fabrication processes. Chemical vapor deposition is commonly used because of the combination of good conformality and high growth rate. With the continuous scaling down of feature sizes and the increase of aspect (depth/width) ratios, it has become a major challenge to maintain the growth conformality and filling efficiency. We developed a numerical model that is capable of simulating the film thickness profiles produced by CVD in trenches and vias with aspect ratios ≥ 10:1. In this model, the precursor transport is described by Knudson diffusion, and the precursor concentration and reaction rate distributions are solved numerically from the continuity equation. The evolution of the deposition profile during a filling process can also be predicted. We employ this model to understand the critical issues associated with CVD on high aspect ratio features. We show that the precursor reaction probability is the key factor that governs the deposition conformality. The model predicts the precursor reaction probability that is required to conformally coat or to completely fill trenches with aspect ratios ranging from 5:1 to 100:1. These simulations are in close agreement with the experimental profiles we have obtained for HfB2 and CrB2 films grown using single-source borohydride precursors. Traditionally the precursor reaction probability is controlled primarily by the growth temperature. We show that if a Langmuir surface reaction mechanisms is operative then the precursor pressure is a far more effective parameter to reduce the reaction probability; i.e. the reaction order vs. pressure must be less than one, corresponding to a surface saturation regime. We have developed a new approach to obtain super-conformal coating (bottom-up filling) of high aspect ratio features, and we have demonstrated proof-of-concept with CrB2 and HfB2 films. In this method, a suppressor species is introduced into the growth system to reduce the surface reaction rate of the precursor; the suppressor pressure is thus a third experimental parameter. We simulate the deposited film profile as a function of the suppressor pressure, precursor pressure, trench aspect ratio, and the relative reaction rates of the suppressor and precursor species. Conditions may be chosen in order to improve the growth conformality of a precursor that is normally too “sticky”. Or conditions may be chosen such that the growth suppression effect gradually weakens along the trench depth, which affords bottom-up filling. We will also discuss methods to optimize the filling process by varying the suppressor pressure in time during film growth.
12:15 PM - D3.7
Thin Polymeric Trench Coatings via Initiated Chemical Vapor Deposition.
Salmaan Baxamusa 1 , Karen Gleason 1 Show Abstract
1 Chemical Engineering, Massachusetts Institute of Technology, Cambridge, MA, Massachusetts, United States
12:30 PM - D3.8
Chemical Vapour Deposition of Boron for Neutron Detector Application
Nirmalendu Deo 1 , Joseph Brewer 1 , Chin Cheung 1 2 , Rebecca Nikolić 3 , Catherine Reinhardt 3 , Tzu-Fang Wang 4 Show Abstract
1 Chemistry, University of Nebraska-Lincoln, Lincoln, Nebraska, United States, 2 Nebraska Center for Materials and Nanoscience, University of Nebraska-Lincoln, Lincoln, Nebraska, United States, 3 Center for Micro and Nano Technology, Lawrence Livermore National Laboratory, Livermore, California, United States, 4 Directorate of Chemistry and Materials Sciences, Lawrence Livermore National Laboratory, Livermore, California, United States
We report the deposition of high purity boron thin films by low pressure chemical vapour deposition (LPCVD) from solid source precursors - 10B and 11B enriched decaborane (B10H14) for neutron detector applications. This boron thin film synthesis is developed and tested for the conformal filling of high aspect-ratio silicon micro-pillar structures for a novel-design neutron detector. Since 10B has high thermal neutron capture cross-section, it is being used as neutron converter elements for detecting thermal neutrons. Compared to conventional boron gaseous precursors such as boron halides and diborane, solid decaborane is a convenient and much less hazardous boron source materials because of its relatively much lower toxicity and pyrophoricity and much higher chemical stability below 150°C. Boron films were deposited in a resistively heated horizontal tube reactor on bare silicon and silicon substrates capped with silicon oxide or silicon nitride with the argon carrier gas and precursor flow controlled by heated mass flow controller. The key process parameters on film purity, roughness, surface structure and growth rate as a function of growth temperature (600-900 °C), different argon flow rates and process pressure (150-450 mtorr) had been examined and correlated with different characterization techniques. Deposition rates can be controlled from 100 to 500 nm/min. X-ray diffraction analysis revealed that boron films were amorphous. Secondary ion mass spectroscopy reveals the very small (20-30-nm) boron diffusion into silicon substrate and 0.5 atomic % of hydrogen detected in the boron films. Both field emission scanning electron microscopy and force microscopy revealed the low surface boron films roughness (1.5-5.5 nm). Preliminary results of conformal filling of 3:1 and higher aspect ratios silicon micro-pillar structures for the fabrication of neutron detector will be discussed.This work was partly performed under the auspices of the U.S. Department of Energy by University of California, Lawrence Livermore National Laboratory under Contract W-7405-Eng-48, UCRL-ABS-225629
D4: Novel Fabrication and Stress Mitigration Strategies
Wednesday PM, April 11, 2007
Room 3003 (Moscone West)
2:30 PM - **D4.1
Porous Thin Film 3-D Nanostructures on Seeded Substrates Grown by Glancing Angle Deposition.
Jeremy Sit 1 Show Abstract
1 Electrical and Computer Engineering, University of Alberta, Edmonton, Alberta, Canada
Glancing angle deposition (GLAD) is an advanced, single-step physical vapour deposition technique that produces porous, nanostructured, microcolumnar thin films with complete control over film morphology. Whereas PVD methods have in general been optimised to produce dense, solid thin films with varying degrees of conformity, the extremely non-conformal GLAD technique takes the opposite approach through the use of collimated vapour flux arriving at non-normal angles (measured with respect to the substrate normal) ranging from oblique to grazing. Deposition onto unheated substates results in limited adatom mobility which is unable to overcome the greatly enhanced atomic shadowing caused by the oblique angle of the incident vapour flux. Using in situ control of the polar and azimuthal directions from which the incident vapour flux arrives, thin films grown by the GLAD technique can be easily sculpted into morphologies such as tilted or vertical columns, or helices or polygonal spirals with precise control over handedness and pitch. GLAD-grown thin films have been demonstrated with a wide range of materials for myriad applications including optical filters, humidity and bio-sensors, and photonic bandgap crystals. Because the formation of the microcolumnar morphology in GLAD-grown thin films depends strongly on the initial nucleation stage of film growth, deposition on non-planar surfaces significantly alters the initial film growth dynamics. Put another way, through careful design of substrate topography, one can achieve an additional dimension of control over the film structure. In short, because of the highly oblique angle of incident vapour flux, substrate surface asperities behave the same way as the initial film nuclei formed on a flat substrate — they produce shawdowed regions on the substrate where the incoming vapour flux cannot directly reach, capturing more of this flux, and hence confine growth of the film to these areas. Thus, by using lithographically patterned dots or lines, the initial growth of the GLAD film structures can be seeded. Using regular arrays of seeds, this technique affords control of periodicity in the plane parallel to the substrate surface; when combined with periodicity in the column morphology in the vertical direction, fully three-dimensional periodic structures can be fabricated in a single deposition step. This paper reviews the basic tenets of the GLAD technique related to deposition onto flat and seeded substrates and presents some of the various structures that can be produced including regular arrays of posts, ribbons, nanofibres, and polygonal spirals for applications such as microfluidics and photonics. Seed size, geometry, layout, and aspect ratio are important considerations and their effect on the resultant GLAD structures will also be discussed.
3:00 PM - D4.2
Full System Model of Magnetron Sputter Chamber - Proof of Principle Study
Chris Walton 1 , George Gilmer 1 , Aaron Wemhoff 1 , Luis Zepeda-Ruiz 1 Show Abstract
1 , Lawrence Livermore National Laboratory, Livermore, California, United States
Lack of detailed process conditions knowledge remains a key challenge in magnetron sputtering, both for chamber design and for process development. Fundamental information such as the pressure and temperature distribution of the sputter gas, and the energies and arrival angles of the sputtered atoms and other energetic species is often missing, or is only estimated from general formulas. However, open-source or low-cost tools are available for modeling most steps of the sputter process, which can give more accurate data from desktop computations than traditional empirical approaches.To get a better understanding of magnetron sputtering, we have collected existing models for the 5 major process steps: the input and distribution of the neutral background gas using Direct Simulation Monte Carlo (DSMC), dynamics of the plasma using Particle In Cell-Monte Carlo Collision (PIC-MCC), impact of ions on the target using molecular dynamics (MD), transport of sputtered atoms to the substrate using DSMC, and growth of the film using hybrid Kinetic Monte Carlo (KMC) and MD methods. Models have been tested against experimental measurements. For example, the gas rarefaction in front of a magnetron observed by Rossnagel and others has been reproduced, and it is associated with a local pressure increase of ~50% which may strongly influence film properties such as stress. Results on energies and arrival angles of sputtered atoms and reflected gas neutrals are applied to the Kinetic Monte Carlo simulation of film growth. Model results and applications to growth of dense Cu and Be films will be presented.Work performed under the auspices of the U. S. Department of Energy by the Lawrence Livermore National Laboratory under Contract No. W-7405-ENG-48.
3:15 PM - D4.3
Novel Vapor Phase Method for Making Ultra Thin Conformal Films.
Sushant Gupta 1 , Rajiv Singh 1 , Arul Chakkaravarthi 1 , Jeff Opalko 2 , Deepika Singh 2 Show Abstract
1 Materials Science & Engineering, University of Florida, Gainesville, Florida, United States, 2 , Sinmat Inc., Gainesville, Florida, United States
3:30 PM - D4.4
Aerosol-Robo-Printing: Writing Ordered Nanostructures on Arbitrary Surfaces with Self-Assembling Inks
Jiebin Pang 1 2 , John Stuecker 2 , Ajay Bhakta 1 , Yingbing Jiang 2 , Joseph Cesarano 2 , Paul Calvert 3 , David Sutton 4 , C. Jeffrey Brinker 1 2 Show Abstract
1 NSF/UNM Center for Micro-Engineered Materials, The University of New Mexico, Albuquerque, New Mexico, United States, 2 Advanced Materials Lab, Sandia National Laboratories, Albuquerque, New Mexico, United States, 3 Department of Materials and Textiles, University of Massachusetts Dartmouth, North Dartmouth, Massachusetts, United States, 4 Strategic Technology Group, Imperial Chemical Industries, Wilton, Redcar, United Kingdom
Evaporation-Induced Self-Assembly (EISA) has been widely employed to synthesize highly ordered mesoporous silica and mesostructured functional organic-inorganic nanocomposites in the forms of thin film and powder. For applications in sensors, micro-optoelectronic devices and microelectromechanical systems, the micropatterning of the mesostructured materials is very important. Several microfabrication techniques, such as surface regulated growth, lithographic micromolding, optical patterning lithography, and high resolution dip-pen nanolithography have been reported for the micropatterning of mesostructured materials. However, some of these lithographic processes are complicated, and most of the micropatterns of mesostructured materials were fabricated on inorganic flat substrates (e.g., silicon, silica, and mica). Developing a facile micropatterning technique that is also friendly to various substrates remains challenging. Our group pioneered the work of micropatterning mesostructured materials through facile micropen lithography and ink-jet printing techniques. In this work, we have demonstrated a new and facile technique to directly pattern mesostructured silica on various substrates (e.g., silicon wafers, glass slides, polymeric transparencies, and curved surfaces) using a robot-controlled aerosol droplet generator and columnated printer system, a so-called Aerosol-Robo-Printer. The aerosol droplets of silica-surfactant sol precursor coalesce into a continuous phase on the substrates and upon evaporation the molecular self-assembly progresses to form highly ordered mesostructures. The pattern profiles can be varied by changing the printing speed, for example, the height and width of printed lines decrease with increasing the printing speed. Highly ordered mesostructures are confirmed by X-ray diffraction and transmission electron microscopy. Increasing the aerosol temperature results in a decrease of the mesostructure ordering, since faster solvent evaporation and enhanced silica condensation at higher temperatures kinetically impede the molecular assembly process. This facile technique using self-assembling inks is of great interest not only in further technical applications of mesostructured materials and other self-assembled nanocomposites but also in developing scientific insights into the EISA process and its extension to directed assembly.
4:30 PM - **D4.6
Through-mask Electrodeposition to Study Intrinsic Stress Evolution During Thin Film Growth
Sean Hearne 1 , Jerrold Floro 2 , Abhinav Bhandari 3 1 , Brian Sheldon 3 Show Abstract
1 , Sandia National Labs., Albuquerque, New Mexico, United States, 2 Department of Material Science and Engineering, U. of Virginia, Charlottesville, Virginia, United States, 3 Division of Engineering, Brown University, Providence, Rhode Island, United States
The sources of stress during polycrystalline thin film growth have been highly debated for over forty years. The debate is the result of the difficultly in separating the multiple stress-creating mechanisms active during stochastically-nucleated thin film growth. To address this, we developed a through-mask electrodeposition technique to constrain the geometry of the islands formed during nucleation, which resulted in a periodic array of identical, nominally cylindrical islands. This was accomplished by electrodepositing Ni through a series of photoresist trenches patterned on a conductive substrate. The resulting islands closely replicated the simplified 2-D geometry typically used in analytical models for stress creation. Additionally, by independently varying the growth rate (over potential) and the island size at the moment of coalescence (trench pitch) we are able to independently study both the kinetically controlled stress creation processes and the geometrically controlled processes. In this presentation, we will discuss our results from this study where we examined the active stress creation mechanisms from the initial moment of island coalescence through the steady state planar growth. Sandia is a multiprogram laboratory operated by Sandia Corporation, a Lockheed Martin Company, for the United States Department of Energy’s National Nuclear Security Administration under contract DE-AC04-94AL85000. The National Science Foundation MRSEC at Brown University (DMR-0520651-IRG1) also supports this research.
5:00 PM - D4.7
Corrugated Diamond Foils by CVD Growth on Patterned Silicon.
Robert Shaw 1 , C. Feigerle 2 , M. Plum 3 , T. Spickermann 4 Show Abstract
1 Chemical Sciences, Oak Ridge National Laboratory, Oak Ridge, Tennessee, United States, 2 Chemistry, Univ. of Tennessee/Knoxville, Knoxville, Tennessee, United States, 3 Spallation Neutron Source, Oak Ridge National Laboratory, Oak Ridge, Tennessee, United States, 4 LANSCE, Los Alamos National Laboratory, Los Alamos, New Mexico, United States
Flat, free-standing carbon foils are required for stripping electrons from a 1 GeV H- beam to yield protons for the Oak Ridge Spallation Neutron Source (SNS). The nominal foil dimensions needed are 12 x 20 mm x 1 micrometer thick; there is an added constraint that they can only be supported along one edge. Traditional foils for similar accelerator applications have been prepared from evaporated carbon, but are sufficiently weak that they must be supported on arrays of thin carbon fibers. This assembly is tedious, and the expected carbon foil lifetime in the 30 mA SNS beam is estimated to be as short as a few tens of hours. We have developed micro- and nanocrystalline plasma-assisted CVD diamond foils for this application with good success. These diamond foils suffer, however, from residual film stress such that when they are released from their silicon growth substrates they curl into an unusable form. This problem has been addressed by first patterning the growth substrate with an asymmetric trapezoidal surface and then growing a conformal diamond film. Upon release from the Si wafer by wet chemical etching, the resulting corrugated foils remain flat.Oxidized, electronic grade 100 Si wafers were scribed and broken to the desired wafer format. Photoresist was spun onto the wafer and exposed as a 1:1 contact print using an optical mask and then developed to form a protective resist pattern on the SiO2 layer. The pattern was transferred to the SiO2 by a brief buffered oxide etch, followed by stripping of the polymer resist. A subsequent anisotropic TMAH etch produced wafer channels with walls inclined at 125 degrees. 1D and 2D corrugation patterns have been explored at 25 to 100 lines/in pitch. Trough depths ranged from 6 to 20 micrometers. The 1D patterns were parallel lines, while the 2D patterns took the form of nested U shapes to retain a flat central foil region. The patterned wafers were abraded by ultrasonic treatment in a mixed slurry of <0.25 and 30 micrometer diamond particles suspended in methanol. Diamond was grown at 50 to 130 Torr pressure at either 1% or 2% methane concentration in hydrogen gas at 1000 to 1300 W microwave power. An excess of argon buffer gas was used to produce nanocrystalline texture.Excellent foil flatness has been attained using this approach. For equivalent corrugation patterns, the nanocrystalline foils reproducibly exhibited slightly better flatness, presumably due to their texture. Microcrystalline foils generally show a slight body twist for the overall free-standing area. We have tested these foils in accelerators in the U.S. and in Japan. At the Proton Storage Ring at Los Alamos a five-month lifetime to failure, corresponding to 820 Coulombs of injected charge, has been recorded for a nanocrystalline foil.** ORNL/SNS is managed by UT-Battelle, LLC, for the U.S. Department of Energy under contract DE-AC05-00OR22725.
5:15 PM - D4.8
Growth of InP on Nanoscale-Patterned Si (100) Substrates by Metalorganic Vapor-Phase Epitaxy
Robyn Woo 1 , Siu Cheng 1 , Li Gao 1 , Robert Hicks 1 Show Abstract
1 Chemical and Biomolecular Engineering, University of California, Los Angeles, Los Angeles, California, United States
III-V materials are promising materials for high speed, low power devices in next-generation integrated circuits. However, a roadblock to implementing these materials in integrated circuit applications is the heteroepitaxial growth of III-V on silicon. InP is the III-V material of interest in our study. The lattice mismatch of 8.1% and thermal mismatch of ~45% introduce a high density of defects in the epitaxial layer. At low temperatures around 250-300 oC by metalorganic vapor phase epitaxy, the growth rate of InP on Si is low while high temperature growth around 600 oC results in 3-D island growth. After implementing two-step growth and thermal cycle annealing, the material still exhibits islands. Here, we show that continuous films of InP can be grown on Si (100) substrates if the substrates are nanometer-scale patterned prior to the growth. The linewidths and the spacings between 1-D trenches are important parameters in the quality of epitaxial film. At the meeting, we will discuss the material quality of the InP layers grown on these nanoscale patterned Si substrates in detail. We will also present results on optical quality of InP grown on Si and how it compares to that of single crystal InP.
5:30 PM - D4.9
Ge Dots Self-assembly on Nanostructured Substrate.
Isabelle Berbezier 1 , Antoine Ronda 1 Show Abstract
1 L2MP, CNRS, Marseille France
Understanding and predicting quantum dots ordering is a central problem for many physical processes. In this work, we investigate the physical mechanisms which govern the self-organization of Ge dots on vicinal Si substrate nanopatterned by growth instabilities. In a first part, we study the formation of large scale nanopatterns by self-organization of growth instabilities. These instabilities develop during MBE growth of Si and Si1-xGex layers on Si substrates. First we investigate the step bunching and faceting instabilities which develop during the growth of Si on vicinal (001) and (111) Si substrates. Evolution of self-organization with time is determined and explained. AFM images are quantitatively analyzed and topographic characteristics are extracted. Critical growth exponents of the instabilities are given and theoretically simulated by kinetic Monte Carlo modeling. Second we investigate the stress-driven kinetic instability which develops during the growth of Si1-xGex on vicinal Si(001). Self-organization of this instability with time, temperature, deposited thickness and strain is analyzed and quantified. Experimental results are compared to theoretical predictions: agreements and discrepancies are discussed. From these experiments we determine the optimal experimental conditions that permit to obtain the best self-organized patterns. At the end of this part, the various typical large scale periodic patterns obtained are presented. In a second part, we analyze Ge dots ordering on these patterns. Depending on the specific features of the patterns developed, Ge dots align either in the valley of the undulations or on their side or on their top. We show that the different Ge dots positioning is due to different driving forces. In particular kinetic surface diffusion arguments, orientation-dependent surface free energy and elastic energy relaxation are used to explain the preferential nucleation of Ge dots in the valley between step bunches, at the intersection between nano-facets and on the top of undulation respectively.
5:45 PM - D4.10
Highly Localized Light Emission From an Organic Device Deposited on an Atomic Force Microscopy Probe Tip Due to Field-assisted Carrier Injection at the Sharp Tip.
Yiying Zhao 1 , Kevin Pipe 2 , Max Shtein 1 Show Abstract
1 Department of Materials Science and Engineering, University of Michigan, Ann Arbor, Michigan, United States, 2 Department of Mechanical Engineering, University of Michigan, Ann Arbor, Michigan, United States
We report on highly localized charge injection and radiative recombination in an organic heterostructure light emitting device deposited onto an atomic force microscopy probe, and the potential application of this mechanism to high resolution scanning optical microscopy. An archetypal (Type A) multi-layer organic light emitting device (OLED) structure consists of organic thin films (on the order of 100nm) sandwiched between an anode and a cathode. Light emission occurs as a consequence of bipolar charge injection from the electrodes and radiative recombination of the charge pairs in the interior of the device. A typical fabrication sequence involves the deposition of a hole transport layer (HTL) onto a substrate coated with a high work function anode, followed by the deposition of an electron transport layer (ETL), followed by vacuum thermal evaporation of a low work function cathode. In contrast, inverted (Type B) OLEDs fabricated on planar substrates using a reversed deposition sequence require specially doped injection layers and often exhibit lower current and light emission efficiency at the same forward bias. It is important to point out that electron injection takes place from the top contact in the Type A device and from the bottom (substrate side) contact in the Type B device.An advantage of organic-based light emitting devices is that high quality organic thin films may be deposited onto a variety of substrates without regard for substrate shape or lattice matching restrictions typically present in inorganic systems. In this work, we compare the optoelectronic performance of Type A and Type B OLEDs deposited on planar and highly non-planar substrates. In particular, we observe that a Type B OLED deposited onto a commercial AFM probe having a sharp pyramidal tip exhibits highly localized current injection and luminescence in the tip region, whereas a Type A device deposited on an AFM probe does not. We attribute this difference to the electric field concentration at the AFM tip, which aids in the injection of electrons in the Type B structure but suppresses electron injection in Type A. This phenomenon is a consequence of the highly non-planar nature of the substrate. This finding also suggests that charge injection and transport in an organic optoelectronic device (or electronic devices generally) can be controlled via a spatially varying built-in electric field caused by non-uniformities or nanostructures in the substrate on which the device is deposited. One of the potential applications of this phenomenon may be an electrically pumped submicron-scale light sources for NSOM-like measurements. The ability to integrate these emitters into an AFM probe tip without additional patterning steps may enable a novel method for simultaneously obtaining topographical and optical images.