Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 53
Frontiers of Engineering: Reports on Leading-Edge Engineering from the 2008 Symposium Molecular and Polymer Nanodevices NIKOLAI ZHITENEV National Institute of Standards and Technology Gaithersburg, Maryland The research and engineering community has been looking intensively for possibilities of extending information-processing technologies into the post-CMOS era. Recently, the Nanotechnology Research Initiative (NRI), a consortium of leading semiconductor companies, has formulated a set of research priorities (Welser et al., 2008) based on an analysis of the ultimate limitations of the present technology and trends in research and development. One of the recommended vectors is devices that operate with state variables different from an electronic charge. One possibility would be a solid-state switch with the computational state defined by the spatial locations of heavy particles, such as ions, atoms, or molecular conformations. The potential advantage of a heavier information carrier can be easily illustrated (Cavin et al., 2006). The scaling of CMOS devices operating with electronic charge will eventually reach their limit when the logic or memory state decays because of electron tunneling under the barriers. For a given barrier height and width limited by material constraints and device size and by the requirement of minimal power dissipation, carriers such as ions or atoms, which are thousands of times heavier than electrons, can provide much greater stability to the compu-
OCR for page 54
Frontiers of Engineering: Reports on Leading-Edge Engineering from the 2008 Symposium tational state. Ironically, the use of heavy carriers is impractical in larger devices because of their limited mobility. However, in a nanometer-sized device, the ion/atom transport can be fast enough for practical applications. Short molecules and macromolecules can be used as active materials for heavier switching devices. Devices built with short molecules have long been considered promising candidates for the post-CMOS era for a number of reasons. First, organic molecules can be extremely small and at the same time exactly reproducible as stand-alone units. In addition, numerous synthetic techniques have been developed, and the variety of organic compounds is enormous. Some well-known approaches to molecular electronics already rely on molecular conformations or oxidation to achieve electronic functionality (e.g., Chen et al., 1999; Collier at al., 2000). However, the reliable fabrication of devices and the assembly of molecules into circuits turn out to be extremely challenging. In this presentation, I will describe some examples from our research that illustrate some of the challenges of fabricating and characterizing molecular devices. Before we began designing a molecular switch or transistor, we tested simpler building blocks in the molecular “tool box,” such as molecular “wires” and molecular “barriers.” The investigation of the electronic properties of molecular devices is intimately related to research on alternative fabrication routes that can be compatible with the new materials. First, the required feature size is often beyond the limits of the best lithography machines. Second, the properties of pristine material can be substantially altered by, for example, exposure to a high-energy electron beam encountered in the e-beam lithography step, etching, or contact deposition. NEW FABRICATION TECHNIQUES In our research, we focused on the noninvasive fabrication of nano- and mesoscale molecular devices and the effects of fabrication on their structural and electronic properties (Zhitenev et al., 2006). We fabricated metal-molecular monolayer-metal junctions using three complementary original techniques that target different fabrication issues. After screening many possible candidates for molecular “wires” and molecular “barriers,” we selected representative molecules capable of forming a dense self-assembled monolayer (SAM) with the most robust structural and electrical properties. The first technique targeted nearly single-molecule devices. The junctions were formed on the surface of the tips to exploit the evaporation of contacts from different angles with an assembly of SAMs in the middle. Device conductance was monitored during the formation of the junction. Devices were studied at multiple stages, from minimally detectable conductance below the conductance level of a single-molecule junction to an approximately single-molecule device to a multimolecule device. The shortcoming of this technique was that it relied
OCR for page 55
Frontiers of Engineering: Reports on Leading-Edge Engineering from the 2008 Symposium on conductance as a single feedback parameter for characterizing the junction formation. The second technique was a planar evolution of the first one. In this case, we used nanoscale stencil masks prefabricated on chip surfaces and angled evaporations to fabricate the molecular junctions with well-defined areas ranging from 30 nm × 30 nm to 1 μm × 2 μm. The third technique, nanotransfer, was designed to avoid the evaporation on top of the SAM and hence to examine the potential damage of evaporation. A column pattern was fabricated on a flexible polymer (PDMS) stamp, and thin metal (gold) was evaporated on the stamp. Functional groups at the top surface of the SAM bonded to the gold when the stamp was brought into contact with the molecular layer. Gold dots with diameters of 20 to 100 nm transferred to the top of the SAM served as contacts to the molecular junction, which was probed by a conducting atomic-force microscope. DISCUSSION Using these three techniques, we were able to examine a variety of fabrication issues, materials properties, and transformations affecting the apparent electrical behavior of molecular devices. We found that it was extremely difficult to fabricate junctions without defects, which can arise from a variety of origins and have a range of effects on device performance. In general, metal electrodes have surface topographical features comparable to or longer than the molecules. Thus the order of molecular assembly was disrupted at the grain boundaries making this location “defective” in a structural and electronic sense. In addition, the nucleation of metal films from the evaporation stream, the resulting surface morphology, the penetration of metal atoms and particles into the molecular layer, and chemical reactions of metals and molecules with oxygen and water were specific to the particular combination of metal and molecular species. For example, junctions with molecular “wires” can appear more resistive than junctions with molecular “barriers” because of the different penetration of gold clusters from the top electrode into the film. If the top electrode is titanium or nickel or another reactive metal, all distinctions between “wires” and “insulators” are lost, because the entire molecular layer is converted into metal carbides or oxycarbides. In general, the electronic levels of molecules were strongly shifted from the Fermi level of metal electrodes, typically by 1 to 5 eV, and the tunneling conductance of such mismatched systems was too low for practical applications. Defects that created electronic states 50 to 200 meV from the Fermi energy level contributed significantly to electronic transport and defined overall behavior. The reliance on precise atomic positions of the device constituents generally failed because the defects took over.
OCR for page 56
Frontiers of Engineering: Reports on Leading-Edge Engineering from the 2008 Symposium NEXT STEPS The question is whether we really need atomic precision to build functional devices. The latest research and engineering results (Scott and Bozano, 2007; Waser and Aono, 2007) have shown that when the overall properties are defined by the self-averaging of multiple imprecise events, the traditional “statistical” approach can be extended to very small devices. For example, we found useful switching functionality in polymer films and the monolayers of macromolecules (Zhitenev et al., 2007) that were just a little bit thicker (5–12 nm) than molecular monolayers (1–2.5 nm). The resistive switching was the result of the shift of electronic levels in the device caused by ionic motion. There is nothing precise about a single-ion position or motion, but the total number of ions in the nanoscale device was large enough to result in statistically reproducible switching. Initially, the devices were nonconducting, but when applied voltage exceeded some threshold level, the devices switched to a conductive state, which was stable at small applied voltage levels. Higher voltage levels of the opposite polarity switched the devices back to a nonconductive state. The switching voltage scaled linearly with the film thickness and depended on the concentration of ionic groups in the film. In addition, ionic groups could be modified in a straightforward way by partly or fully replacing protons with other ions. Because many of these substitutions were reversible, we were able to examine multiple chemical compositions with the same mesoscopic devices. When we did so, we found that the chemical modifications had major effects on switching behavior. A simple physical model that captures the most essential experimental finding is described below. In the “off” state, all molecular energy levels are a few electrovolts from the electrode Fermi level, and switching the electric field is strong enough to break the ionic bonds. The ion separation acts as internal “chemical gating,” shifting some energy levels into better alignment with the Fermi energy of the electrodes. These electronic states, with energy levels of 100–300 meV from the Fermi level, form the conducting channels. A strong electric field of the opposite polarity pushes the separated cations back, facilitating the recombination with anions at the polymer backbone and eliminating the conductive electronic levels. The electric field required for ion separation depends on the size and properties of the ions. For example, larger monovalent cations have smaller bond strength. Thus the devices can be turned on at smaller threshold voltage. With multivalent ions forming chemical bonds to two or more anion groups, the threshold voltage is significantly increased. Polymer switches are just one example of material systems that display resistive bistability. There are many other candidates based on various organic and inorganic compositions (Scott and Bozano, 2007; Waser and Aono, 2007). For all of these materials, there is a common element in switching behavior. The
OCR for page 57
Frontiers of Engineering: Reports on Leading-Edge Engineering from the 2008 Symposium switching functionality is caused by the movement or displacement of heavy particles, such as ions or atoms, over distances ranging from an elementary cell to the size of an entire device. The variety of materials capable of displaying switching leads us to believe that devices based on atom/ion motion can eventually be used in practical circuits. CONCLUSION The use of switching in memory and storage devices has been the main driver for the development of switches by most of the major semiconductor companies. However, if the switching phenomenon can be reliably engineered in devices of sufficiently small size, this will lead to the emergence of new hybrid logic circuits based on novel architectural concepts (Strukov and Likharev, 2007). Some of these concepts mimic the “architecture” and the well-developed connectivity of the human brain, an integral combination of memory, connectivity, and computational elements. REFERENCES Cavin, R.K., V.V. Zhirnov, D.J.C. Herr, A. Avila, and J. Hutchby. 2006. Research directions and challenges in nanoelectronics. Journal of Nanoparticle Research 8(6): 841–858. Chen, J., M.A. Reed, A.M. Rawlett, and J.M. Tour. 1999. Large on-off ratios and negative differential resistance in a molecular electronic device. Science 286(5444): 1550–1552. Collier, C.P., G. Mattersteig, E.W. Wong, Y. Luo, K. Beverly, J. Sampaio, F.M. Raymo, J.F. Stoddart, and J.R. Heath. 2000. A catenane-based solid state electronically reconfigurable switch. Science 289(5482): 1172–1175. Scott, J.C., and L.D. Bozano. 2007. Nonvolatile memory elements based on organic materials. Advanced Materials 19(11): 1452–1463. Strukov, D.B., and K.K. Likharev. 2007. Defect-tolerant architectures for nanoelectronic crossbar memories. Journal of Nanoscience and Nanotechnology 7(1): 151–167. Waser, R., and M. Aono. 2007. Nanoionics-based resistive switching memories. Nature Materials 6(11): 833–840. Welser, J.J., G.I. Bourianoff, V.V. Zhirnov, and R.K. Cavin. 2008. The quest for the next information processing technology. Journal of Nanoparticle Research 10(1): 1–10. Zhitenev, N.B., W.R. Jiang, A. Erbe, Z. Bao, E. Garfunkel, D.M. Tennant, and R.A. Cirelli. 2006. Control of topography, stress and diffusion at molecule-metal interfaces. Nanotechnology 17(5): 1272. Zhitenev, N.B., A. Sidorenko, D.M. Tennant, and R.A. Cirelli. 2007. Chemical modification of the electronic conducting states in polymer nanodevices. Nature Nanotechnology 2(4): 237–242.
OCR for page 58
Frontiers of Engineering: Reports on Leading-Edge Engineering from the 2008 Symposium This page intentionally left blank.