Spring 2003-2004 Schedule
Performance Limitations of Devices and Interconnects and Possible Alternatives for Nanoelectronics
April 9, 2004; 3:30 pm
Presented by: Prof. Krishna C. Saraswat Department of Electrical Engineering, Stanford University, Stanford, CA 94305
For over three decades, there has been a quadrupling of transistor density and a doubling of electrical performance every 2 to 3 years. Si transistor technology, in particular CMOS has played a pivotal role in this. It is believed that continued scaling will take the industry down to the 35-nm technology node, at the limit of the "long-term" range of the International Technology Roadmap for Semiconductors (ITRS). However, it is also well accepted that this long-term range of the 70-nm to 35-nm nodes remains solidly in the "no-known solution" category. The difficulty in scaling the conventional MOSFET makes it prudent to search for alternative device structures. This will require new structural, material and fabrication technology solutions that are generally compatible with current and forecasted installed Si manufacturing. In addition, new and revolutionary device concepts need to be discovered and evolved. These can be split into two categories: one is the continued used of silicon FET-type devices but with additional materials, e.g., Ge and innovative structural aspects that deviate from the classical planar/bulk MOSFET, e.g., double gate MOSFET. The second category is a set of potentially entirely different information processing and transmission devices from the transistor as we know it, e.g. silicon-based quantum-effect devices, nanotube electronics and molecular and organic semiconductor electronics.
Continuous scaling of VLSI circuits can pose significant problems for interconnects, especially for those responsible for long distance communication on a high performance chip. Our modeling predicts that the situation is worse than anticipated in the ITRS, which assumes that the resistivity of copper will not change appreciably with scaling in the future. We show that resistance of interconnect wires in light of scaling induced increase in electron surface scattering, fractional cross section area occupied by the high resistivity barrier and realistic interconnect operation temperature will lead to a significant rise in the effective resistivity of Cu. As a result both power and delay of these interconnects is likely to rise significantly in the future. In the light of various metal interconnect limitations, alternate solutions need to be pursued. We focus on two such solutions, optical interconnects and three-dimensional (3-D) ICs with multiple active Si layers.
X-ray spectroscopy studies of interfaces
April 16, 2004; 3:30 pm
Presented by: Prof. Anders Nilsson Stanford Synchrotron Radiation Laboratory
Water is the key compound for our existence on this planet and it is involved in nearly all biological, geological and chemical processes. The ability to break and reform hydrogen bonds makes these systems extremely flexible. Hydrogen bonding at surfaces and interfaces represent another level of complexity compared with bulk water. The understanding of water surfaces is essential for many processes of importance in hydrogen technology, catalysis, electrochemistry, biomaterials, molecular environmental science and atmospheric chemistry. X-ray spectroscopy using synchrotron radiation probes the local electronic structure around specific atomic sites. It will be demonstrated how we can use x-ray spectroscopy to both probe the nature of chemical bonding of water to metal surfaces and hydrogen bonding between water molecules.
Capturing singular sets in solids: From crystallographic slip to faults in the earth's crust
April 21, 2004; 4:15pm
Presented by: Prof. Michael Ortiz Aeronautics and Mechanical Engineering Caltech
When solids are deformed beyond their initial elastic range, a competition often arises between localized and de-localized deformation process. Examples of this competition are: crystallographic slip on discrete slip planes in ductile crystals; localization of plastic deformation into adiabatic shear bands; decohesion, cleavage fracture and fragmentation; and faults in the earth crust. Localization processes are ubiquitous because they relate to fundamental aspects (convexity vs concavity) of the energetics of solids. From a computational perspective, the localized deformations often cannot be resolved by the volume grid and must be treated as sub-grid phenomena. Mathematically, this corresponds to idealizing localized deformations as occurring on a low dimensionality 'singular set.' I will review a number of developments in computational mechanics and mathematical analysis aimed at capturing such singular sets and shedding light on the properties of the corresponding solutions. These developments include: phase-field models of dislocation dynamics; localization elements for analyzing strain localization; cohesive elements for the analysis of fracture and fragmentation; and models of distributed faulting.
Evaporation-Induced Self-Assembly of Porous and Composite Thin Film Nanostructures
April 30, 2004; 3:30pm
Presented by: Prof. C. Jeffrey Brinker Dept. of Chemical and Nuclear Engineering and Chemistry, University of New Mexico, Albuquerque, NM; Dept. of Materials Chemistry at Sandia National Lab, NM.
Nature combines hard and soft materials often in hierarchical architectures to get synergistic, optimized properties and combinations of properties with proven, complex functionalities. Emulating such natural material designs in robust engineering materials using efficient processing approaches amenable to manufacturing represents a fundamental challenge to materials scientists and engineers. Currently there is considerable interest in evaporation-driven self-assembly as a means to create porous and composite thin film nanostructures using simple commercial procedures like dip or spin-coating and ink-jet printing. This presentation will first review recent progress on evaporation-induced silica/surfactant self-assembly (EISA) to prepare porous thin film nanostructures of interest for membranes, sensors, and low K dielectrics. Starting with a homogenous solution of surfactant plus hydrophilic oligosilicic acid precursors, solvent evaporation concentrates the depositing film in precursors and surfactant inducing micelle self-assembly and further self-organization into thin film silica/surfactant mesophases. Exploiting the steady, continuous nature of dip-coating, it is possible to spatially resolve the complete evaporation-induced self-assembly pathway (in the coating direction) and interrogate it using spectroscopy and/or grazing incidence SAXS. I will then discuss surfactant self-assembly as a means to organize simultaneously hydrophilic and hydrophobic precursors into hybrid (organic/silica or metal/silica) nanocomposites that are optically or chemically polymerizable, patternable, or adjustable. For example in a recent paper we demonstrate self-assembly and integration of robust, adjustable 3D nanocrystalline arrays. (Science April 23, 2004). Biocompatible self-assembly, using phospholipids as the structure-directing agents, allows cell immobilization in a robust self-contained, self-sustaining environment of interest for stand-alone cell-based sensors. However, we observe that cells co-opt the EISA process, altering significantly the self-assembly pathway and creating a unique bio/nano interface. As a new direction in self-assembly, we have exploited mechanically-based re-assembly to create superhydrophobic, fractal silica surfaces mimicking those of the Lotus leaf and desert beetle. These surfaces are self-cleaning and fundamentally affect flow, making them of general interest for fluidic-based microsystems.
Partitioning the Sun's Spectrum with Novel Materials in High-Efficiency III-V Multijunction Solar Cells
May 7, 2004; 3:30pm
Presented by: Dr. Richard King Spectolabs, Inc. Sylmar, CA
Sunlight is an energy resource with a wide geographic distribution, making it easily accessible for solar electricity generation in many communities, as well as in space. However, it is also distributed over a wide range of wavelengths, complicating efficient photovoltaic energy conversion. Multijunction solar cells divide the solar spectrum into narrower wavelength bands, allowing more efficient conversion by each subcell with bandgap energy tuned to the photon energies in the corresponding band. Many semiconductor systems, such as III-V, III-N, II-VI, I-III-VI, and organic semiconductors have bandgaps that are variable over a wide range of energies relevant for photovoltaic conversion. The GaInP/ GaInAs/ Ge 3-junction cell is a multijunction cell (MJ) configuration combining III-V and group-IV subcells, that has achieved record efficiencies under both the terrestrial and space solar spectra. Efficiencies well over 40% are possible in principle if a different combination of subcell bandgaps can be used to partition the solar spectrum. The sun's surface temperature and resulting shape of its spectrum result higher theoretical MJ cell efficiencies for higher top cell bandgap, lower middle cell bandgap, and for MJ cells which employ a ~1-eV subcell, such as GaInNAs, above the bottom 0.67-eV Ge subcell. This talk examines unconventional semiconductor materials that can be used to achieve the desired combination of bandgaps in new types of high-efficiency MJ solar cells. Threading dislocations and resulting lower minority-carrier lifetime in lattice-mismatched materials are discussed, out to compositions of 35%-In GaInAs and 82%-In GaInP, with a lattice constant 2.4% larger than the Ge substrate. The ability of group-III sublattice ordering to change the bandgap of GaInP without varying its composition or lattice constant is examined. The dilute (~2%) nitrogen alloy GaInNAs, one of the few ~1-eV semiconductors that is lattice-matched to Ge, provides a way to convert excess photogenerated current density that would otherwise be wasted in the Ge subcell. The unusual band structure and high density of as-grown defects in GaInNAs result in low mobilities and lifetime, and have made such subcells very challenging to use in MJ cells without limiting the current through the series-interconnected subcell stack. This talk discusses results measured on the first 6-junction cells built, which make use of a GaInNAs 1.1-eV subcell, with open-circuit voltages over 5.1 V. Record efficiency 3-junction III-V cell results are discussed for space-optimized cells (30.1%, AM0), and terrestrial concentrator cells (36.9%, 309 suns).
Materials Challenges in Designs of Gigahertz Silicon Optical Modulators Based on a Metal-Oxide Semiconductor Capacitor
May 14, 2004; 3:30pm
Presented by: Dr. Dean Samara-Rubio Intel Corporation Santa Clara, CA
It is well-known that low-loss silicon waveguides can be fabricated on SOI substrates. SOI waveguide devices such as arrayed waveguide gratings (AWG), Bragg filters, variable optical attenuators (VOA), photodiodes, and others have been developed for telecom applications. The relative ease of manufacturing and low cost of Si-based devices has fuelled the drive to increase the variety and quality of optical functions which can be monolithically integrated with SOI waveguides. However, two key functions are still lacking: high-speed data encoding (optical modulation), and efficient lasing (light generation/amplification).
Here we describe an SOI waveguide optical modulator which operates at speeds in excess of 1GHz. This is a significant milestone since the fastest SOI-based optical modulator measurement found in literature is ~20 MHz. The device reported is based on free-carrier plasma dispersion in an accumulation-mode metal?oxide?semiconductor (MOS) capacitor. Incorporating the MOS stack into the waveguide is a departure from the previous works which relied on carrier injection across forward-biased diode junction(s). The primary advantage of the MOS capacitor operating in accumulation bias is the elimination of the minority carrier lifetime from the optical response of the modulator. The conversion to an MOS structure, however, poses some materials and design challenges that will be discussed, including the need for highly-doped silicon and poly-silicon in the optical path, changes in optical characteristics due to the presence of a low-index gate layer, and a constraint on the available free-carrier density due to gate reliability limits.
Finally, we note that at speeds beyond 1GHz, improved electrical-to-optical efficiency (charge efficiency) of the device will be needed to manage the electrical driver power dissipation. We will outline the techniques for improved efficiency and highlight the anticipated challenges for the materials and processes.
Printed polymer electronics: from material properties to display backplanes
May 21, 2004; 3:30pm
Presented by: Dr. Alberto Salleo Palo Alto Research Center
There is great interest in using organic semiconductors as active materials in thin-film transistors (TFTs) for active-matrix display backplanes. The benefits introduced by the use of polymeric semiconductors include low temperature processing and deposition from a liquid phase. As a consequence, device arrays can be jet printed on plastic substrates to enable low-cost, large-area electronics for displays, sensors, and evolving technologies such as electric paper. The use of polymer semiconductors must however rely on a better understanding of the properties of these materials and their impact on device operation. The operation of TFTs is governed by the material within ~1nm of the semiconductor/gate dielectric interface. Hence, functionalizing the dielectric with a self-assembled monolayer (SAM) leads to a performance improvement of up to three orders of magnitude in poly(thiophene) TFTs. Device performance has increased steadily in the past years, however there remains the question as to what is the highest mobility that can be achieved in polymer TFTs. In order to address this question, I will show that measurements of poly(thiophene) mobility as a function of temperature agree well with a multiple trapping and release (MTR) model. The structural disorder in the material generates a shallow donor-like trap distribution at the edge of the valence band; the width of this band is a function of the amount of disorder in the film. According to the MTR model, the maximum mobility of the mobile charges in the film is estimated to be between 1 and 4 cm^2/V.s. Finally, I will show examples of non-ideal behavior, such as a slow output current decrease under DC gate bias. Recent studies of these bias-stress effects in regio-regular poly(thiophene) TFTs indicated that hole pairs form self-trapped bipolarons which are responsible for at least one component of the bias-stress effect. A semi-quantitative kinetic model of bipolaron formation agrees well with the experimental results and allows to estimate the binding energy of the bipolaron (~120 meV).
Efficiency limits of organic solar cells: how do we get there?
May 28, 2004; 3:30pm
Presented by: Prof. Peter Peumans Electrical Engineering Dept. Stanford University, CA
Thin-film organic solar cells have attracted attention over the last decade because of their potential to substantially reduce the cost of solar electricity by lowering the cost of the substrate, materials, processing and installation. In addition, organic materials allow very thin and flexible cells to be designed, leading to entirely new device concepts that may revolutionize the way we harvest solar energy, such as roll-up solar cells and solar textile fiber. The power conversion efficiency of organic solar cells has increased steadily since the demonstration of the donor-acceptor cell architecture in 1986 to the current state of the art of 3-4%, but remains far below that achieved in amorphous (~12%) and crystalline silicon (~24%) cells. In this talk, I will analyze the efficiency limits of organic solar cells and compare the current state-of-the-art with these limits. It will be shown that the voltage efficiency and fill factor of a single junction organic cell are correlated, limiting the maximum achievable efficiency of a single junction cell. To further improve the power conversion efficiency, we have developed multijunction architectures that allow several cells to be stacked in a series connection. These devices rely on metal nanoclusters as effective recombination layers between adjacent cells. The optimization and fabrication of multijunction organic solar cells with wide spectral coverage is discussed.