as they are stimulated by the addition or removal of a sin-or a domino - PDF document

as they are stimulated by the addition or removal of a sin-or a dominolike arrangement of molecules inlar-mechanical cascade that ripples through the whole as-sembly in a predetermined fashion.which the presence or absence of a single atom dominatesthe electrical conductivity,recognition process is used to open up new conductivity 2003 American Institute of Physics, S-0031-9228-0305-020-8 May 2003 Physics Today Jim Heath Improvements in our understanding of how moleculestransport charge,and how they interface to the macro-scopic world,are fueling new devices and applications.James R.Heath and Mark A.Ratner modify electronic behavior, providing both switching andDynamical stereochemistry. Many molecules have mul-Synthetic tailorability. By choice of composition andgeometry, one can extensively vary a molecule’s transport,erties into a molecule. Akey challenge in molecular elec-Apicture of electron transport through molecular de-vices is emerging, and it couldn’t be more different fromohmic: For a given wire diameter, longer wires have pro-four types of molecular-electronic junctions are repre-other as an electron acceptor. The electrodes are bridgedmuch richer for the electron donor-bridge-electron accep-tor (DBA) molecular junction of figure 3b. DBAcomplexessolid-state molecular junctions. In DBAcomplexes, theponent that has molecular orbitals of differing energy. Instate before tunneling to the left electrode. Alternatively,processes will occur, and it is their relative rates that de-DBAjunction.Athird possibility is that an electron fromis quite closely related to ohmic charge flow. May 2003 Physics Todayhttp://www.physicstoday.org 01.02.0mm ab Figure 1. Molecular electronics devices.(a)molecular circuits, fabricated by a combination of soft-imprinting techniques for the wires and chemical An atomic force micrograph of one of the circuits,which could be used either as a random access memory or as a combination logic and memory circuit. Themolecules used in this circuit are bistable [2]rotaxanes. (After Y. Chen et al., ref. 13. Courtesy of StanWilliams, Hewlett–Packard Co.) DBAjunctions illustrate some of the beauty and rich-ness of molecular electronics. From a chemist’s perspec-tive, the diversity of conduction mechanisms representsjunctions through synthetic modification. The observedconduction in DBAmolecular junctions usually differsradically from that in traditional ohmic wires and canstructures. Key factors include a dependence on the ratesof intramolecular electron transfer between the donor anda DBAjunction can vary with the sign of the applied volt-age; such junctions represent a molecular approach to-nisms through a DBAmolecule can also be affected by thepler energy level system than DBAjunctions, and have be- http://www.physicstoday.orgMay 2003 Physics Today 100 nm LLLLRRRRabcd DBAwxyzHSSHHSSH lEB NNNNN +N++N+N N SSS each approximately 5nm in di-ameter. The lattice constant is 15nm. Certain materials pa-nanoscale dimensions. At this size scale, however, chemical The top panels depict molecules with various localized, low-energymolecular orbitals (colored dots) bridging two electrodes L(left) and R (right). In the middle panels, the black lines are unpeturbed electronic energy levels; the red lines indicate energy levels under an applied field. The bottom panels depict repre-Alinear chain, or alkane. Adonor-bridge-acceptor (DBA) molecule, with a distance between the donor and acceptor and an energy difference Amolecular quantumAn or-a [2]rotaxane, which displays a diverse set of localized molecular sites along the extended chain. Two of those sites (red andgreen) provide positions on which the sliding rectangular unit (blue) can stably sit. Asecond example of a complex moleculebridging the electrodes might be a short DNAchain. electrode measurements. An equally important advantagefew meV. Such resolution allows, for example, measure-Two recent papers reported on a unique quantum ef-Electrode effectsstructing molecular-electronic devices are based on prac-unpredictably modify the molecular component. Ideally,tions. However, the current state of the art for the theoryecule and electrode. Consequently, at zero applied biastially or fully mask the molecule’s electronic signature, in-son—and others, including stability, reproducibility, andVery little theory exists that can adequately predicthow the molecular orbitals’energy levels will align with theical and demands both theoretical and experimental study.Arelated consideration involves how the chemical nature May 2003 Physics Todayhttp://www.physicstoday.org TAGE(mV)€€ GATE VOLTAGE(V) V(/)V(0) organic bridge (denoted as the two white spheres separatedaround 1V, at which point Wire Junctionthat charge can move through such a structure by elasticMathematically, the Landauer formula isis the electron charge, Planck’s constant, and (12.8kis the Green’s function that characterizes electronspectively. Electronic structure theory permits actual eh2RMLM. ehT
22. describe the molecule’s electronic structure and the mole-cule–electrode interfaces (see box 1 on page 46). However,trode will likely modify the molecule’s electron density ins electron density inmight appear large and complex, but it is actually smallical, and electronic properties that have been built into it.DNAoligomers represent perhaps the best-studied ex-perimental example of this category. In addition to its bi-of molecular nanostructures, the DNAmolecule is of in-Intramolecular electron transfer rates in DNAhavea large mechanistic diversity. In general, for very short-ing can occur. For transfer over more than six or seven basecillators that account for the chemical environment. Antance has been seen directly in DNAmolecules folded intomechanistic palette observed for DNAcharge transfer ispossibilities, DNAmight act as a paradigm for electronElectrical transport in DNAmolecular junctions istural dynamics and disorder, geometric reorganization,and sample preparation all add to the intricacies. As a re-nals have stated that DNAacts as an insulator, a semi-conductor, a metal, and a superconductor.Most probably, transport in DNAjunctions will showby localized hole hopping between the low-energy guanine–Because the bandgap is large, DNAappears uncolored andpanied by molecular distortion; Anderson-type charge electrons localized on GC and adenosine–thymine (AT)pairs; structural reorganization; counter-ion motion; andsolvent dynamics. The available data are, more or less, con-orators that DNAis a wide-bandgap semiconductorcan exhibit activated transport for relatively short dis-insulator at distances exceeding 20 nm. The complexity andrichness of DNAjunction behavior typify the challenge thatthe molecular electronics community faces in predicting http://www.physicstoday.orgMay 2003 Physics Today HOLE TRANSFER EFFICIENCY10210100 LENGTH OF BRIDGE (Å)6050403020100 CoherenttunnelingIncoherent hopping Figure 5. DNAshows a competitionbetween different chargeangles and circles) and theoretical (solid line) results for therelative rate of hole transfer between guanine–cytosine (GC)base pairs on DNAoligomers. The theoretical model incor-hopping between GC pairs separated by a bridge of severaladenosine–thymine pairs, which have higher energy. Coher-characteristic exponential decay. Incoherent hopping domi- circuitry has advanced quickly. The proposed circuit ar-dard lithography; and fabrication simplicity.those considerations is the crossbar,ing individual molecular or molecular-scale devices sand-Arapidly developing area of architectural researchcircuit components—a molecular-electronic random accessmemory. This particular circuit satisfies all five of the keyThe fundamental challenges of realizing a true molecular-molecular devices. Additional challenges involve findingthrough molecular assembly. Controlling the properties ofmolecule–electrode interfaces and constructing molecular- May 2003 Physics Todayhttp://www.physicstoday.org ne of the most attractive architectures for designing molecular-electronics circuits for computational applications andinterfacing them to the macroscopic world is the crossbar. The general concept is shown on the left, where a sort ofpatchwork quilt of logic, memory, and signal routing circuits is laid out. The simplest of these circuits—and one that hasbeen experimentally demonstrated—is a memory circuit.The memory, shown on the right, consists of two major components. The central crossbar—the crossing of 16 verticaland 16 horizontal black wires—constitutes a 256-bit memory circuit. Bistable molecular switches are sandwiched at theEach set of the larger blue wires is arranged into what is called a binary tree multiplexer. The multiplexers here adoptsome interesting architectural variations that allow them to bridge from the micron or submicron scale of the blue wires topairs for each multiplexer.address (0110, for example) isas a four-input AND gate soindicate how one wire (red) is selected by each multiplexer.tacts is much larger than the pitch of the nanowires; thatlarger separation greatly reduces the fabrication demands. Note also that the frequencies of the patterns of connections areis not important. Those two characteristics allow the architecture to bridge the micron or submicron length scales of lithography to the nanometer length scales of molecular electronics and chemical assembly. LogicLogicMemoryRouting and interconnects011 lenges. Voltage-gated, single-molecule devices may emergeeventually link experiment and theory. Binary tree multi-tested against any of the “ilities”: reliability, temperaturestability, and the like.lecular-electronic integrated circuitry.We are grateful to our research groups and colleagues forteachings and ideas, and to NSF, Semiconductor ResearchCorp, and the US Department of Defense (including the Office1.C.P. Collier et al., , 1172 (2000); Y. Luo et al.,, 519 (2002).2.A.J. Heinrich et al., 3.J. Park et al., , 722 (2002); W. Liang et al., 4.D.N. Boon, J.K. Barton, Curr. Opin. Struct. Biol.F.D. Lewis et al., , 11280 (2002).5.N. Melosh et al., , 112 (2003).6.P. Packan, 7.A. Nitzan, Annu. Rev. Phys. Chem., 681 (2001); V. Mujica,M.A. Ratner, in Technology, W.A. Goddard III et al., eds., CRC Press, BocaRaton, Fla. (2002); C. Joachim, J.K. Gimzewski, A. Aviram,8.A. Aviram, M.A. Ratner, , 277 (1974);R.M. Metzger et al., 9.M.A. Reed et al., 10.Y.A. Berlin, A.L. Burin, M.A. Ratner, 11.D. Porath, A. Bezryadin, S. de Vries, C. Dekker, 12.J.R. Heath, P.J. Kuekes, G. Snider, R.S. Williams, 13.A. Bachtold, P. Hadley, T. Nakanishi, C. Dekker, 1317 (2001); Y. Huang, , 1313 (2001); P. Avouris et, 6 (2002); Y. Chen et al., 14.A. DeHon, in Proc. First Workshop on Non-Silicon Computa-, available at http://www.cs.caltech.edu/research/ic/15.P.J. Kuekes, R.S. Williams, “Demultiplexer for a MolecularWire Crossbar Network,” US Patent 6,256,767 (3 July 2001). May 2003 Physics Today Circle number 20 on Reader Service Card