atomic, molecular, and optical physics workshopastro1.panet.utoledo.edu/~ljc/amopfr.pdf · atomic,...

Atomic, Molecular, andOptical PhysicsWorkshop

Final Report

of the workshop heldSeptember 21-24, 1997inChantilly, Virginia

Sponsored by:

The Division of Chemical Sciences

Office of Basic Energy Sciences

Office of Energy Research

U.S. Department of EnergyWashington, D.C.

Table of Contents ii

Table of Contents

This document was produced under contract number DE-AC05-76OR00033 between the U.S. Department of Energy and Oak Ridge Associated Universities.

Preface ...................................................................................................................................................................... iii

FORWARD ................................................................................................................................................................ v

Panel A: Interaction of Atoms and Molecules with Photons - Low Field ................................................................ 7

Panel B: Interactions of Atoms and Molecules with Photon - High Field ............................................................. 18

Panel C: Surface Interactions With Photons, Electrons, Ions, Atoms, and Molecules ............................................ 31

Panel D: Theory of Structure and Dynamics ........................................................................................................... 45

Panel E: Nano- and Mesoscopic Structures............................................................................................................. 57

iii Preface

This document contains the final reports from the five panels that comprised a Workshop held to explore futuredirections, scientific impacts and technological connections of research in Atomic, Molecular and Optical Phys-ics. This workshop was sponsored by the Department of Energy, Office of Basic Energy Sciences, ChemicalSciences Division and was held at the Westfields International Conference Center in Chantilly, Virginia onSeptember 21-24, 1997. The workshop was chaired by Lloyd Armstrong, Jr., University of Southern Californiaand the five panels focused on the following topics:

Panel A: Interactions of Atoms and Molecules with Photons - Low FieldDaniel Kleppner (Massachusetts Institute of Technology), chair

Panel B: Interactions of Atoms and Molecules with Photons - High FieldPhil Bucksbaum (University of Michigan), chair

Panel C: Surface Interactions with Photons, Electrons, Ions, Atoms and MoleculesJ. Wayne Rabalais (University of Houston), chair

Panel D: Theory of Structure and DynamicsChris Greene (University of Colorado), chair

Panel E: Nano- and Mesocopic StructuresPaul Alivisatos (Lawrence Berkeley National Laboratory), chair

The choice of focus areas reflects areas of significant interest to DOE/BES but is clearly not intended to span allfields encompassed by the designation of atomic, molecular and optical physics, nor even all areas that would beconsidered for review and funding under DOE’s AMOP program. In a similar vein, not all research that might besuggested under these topics in this report would be appropriate for consideration by DOE’s AMOP program. Theworkshop format included overview presentations from each of the panel chairs, followed by an intensive seriesof panel discussion sessions held over a two-day period. The panels were comprised of scientists from the U. S.and abroad, many of whom are not supported by DOE’s AMOP Program. This workshop was held in lieu of thecustomary “Contractors Meeting” held annually for those scientists currently receiving support from the DOEAMOP Program. The current DOE principal investigators attended the workshop, presented posters on theircurrent research, and engaged in discussions with the panel members. Thus, the reports contained in this docu-ment reflect their active participation.

The workshop and this report are intended to stimulate and provoke the thinking of those interested in research inAMOP. Anyone who would like to submit a proposal (Grant Application) for consideration and review can do soat any time. This document as well as the annual Summaries of Research in the Chemical Sciences should beuseful reference materials to consider. Both are available either as hard copy or on our website at:http://www.er.doe.gov/production/bes/chm/chmhome.html

PREFACE

Preface iv

The web version of the FY 1997 summaries of research can be easily searched by several criteria including bothscientific discipline and relevant technologies. Those considering submitting a proposal are encouraged toprepare a brief preproposal prior to preparing and submitting a full proposal. Preproposals can be submitted anddiscussed with staff in the Chemical Sciences Division. All relevant information, including phone numbers and e-mail addresses, for all Division staff is available on our website and in the hardcopy version of the researchsummaries. Preproposals should be addressed to the Division Director with copies to the appropriate TeamLeader and Program Manager. This will ensure full and timely consideration.

Patricia DehmerAssociate Director of Energy Research for theOffice of Basic Energy SciencesU. S. Department of Energy

v Forward

FORWARD

Recent years have seen a virtual explosion of activity and advances in the fields of atomic, molecular andoptical physics. Development of new instrumentation and new experimental techniques has proceeded at anunusually rapid rate, leading to discovery of many new phenomena and much improved characterization ofpreviously known phenomena. New theoretical approaches have enabled higher levels of understanding ofold problems, helped to provide explanations for newly discovered phenomena, and pointed out directionsfor new exploration.

The Department of Energy Office of Basic Energy Sciences held a workshop on Atomic, Molecular, andOptical (AMO) Physics on September 21-24, 1997, to review advances in selected areas of AMO physicsand to consider the impact of these changes on the closely allied fields of surface interactions andnanostructures. Emphasis was placed on identifying future directions in all of these areas that are ripe forexploitation. The papers that follow describe the conclusions of the AMO panels that reviewed high-fieldand low-field regimes and the theory of structure and dynamics, as well as of the panels that looked atsurface interactions and nanostructures and their relations with AMO physics.

One of the central themes running through all of these reports is that many of these new developments enableor should in the future enable a great expansion in our ability to control broad classes of physical phenomena.Control of novel photon and atom states, of quantum dynamics of collisions and chemical bonds, of atoms onsurfaces, of chaotic behavior in nonseparable dynamical systems, of size, shape and properties ofnanostructures is increasingly possible. Control of these and other similar phenomena promises to open upnew and exciting fields of research, and to lead to the creation of powerful new experimental and theoreticaltechniques for further exploration. Achieving this increased level of control and of the understanding ofthese complex systems that this control implies could enable very significant advances in many importantareas of research. For example, understanding and control of the quantum dynamics of collisions andchemical bonds is fundamental to the control and manipulation of chemical reactions.

Another important interconnecting theme is that of the continuing rapid evolution of probes, in particularthose that utilize light. Developments in lasers are leading rapidly to ever higher intensity, ultrafast pulses ofdefined shape, shorter wavelength, greater stability and narrower lines. Third generation synchrotron lightsources are providing unprecedented resolution and sensitivity in the high-energy photon range. Newapproaches to the production of photon beams having nonclassical properties have the potential to enablespectroscopic sensitivity beyond the standard quantum limit. These and other new probes enable importantnew experiments not only in AMO research, but in many other fields supported by the Office of BasicEnergy Sciences as well.

Several of the reports discuss recently created novel states of matter or matter and field. The ultimate utilityand application of these states is unknown, but many seem to have great potential for significant impact.Bose Einstein condensates (BEC) are one of these. Already, researchers have been able to use BEC to create“atom lasers” that display phase coherence between beams of atoms, raising the possibility of, for example,

Forward vi

greatly increased capabilities in lithography. In other work, new states are being created in which matter andfield form systems with unusual properties. Among these at one end of the energy spectrum are entangledstates, created in micromaser cavities with a small number of photons and atoms; at the other end are states inwhich external field strengths are comparable to the strength of the binding fields of the matter.

The papers on nanostructures and surface physics clearly show the very significant impact of existing AMOtechnologies and theories on these fields. In addition, as workers in these fields look to the future, they findnumerous examples where developing AMO technologies should be of considerable importance in enablingthe fields to move ahead. This theme of AMO as an enabling science is found throughout all of the reports,with each paper discussing areas outside of AMO where new technologies, instrumentation and theories beingdeveloped to advance the physics of AMO might find applications. Numerous potential important enablingapplications of AMO technologies to other areas of interest to DOE and in particular BES are indicatedthroughout the reports.

Since the emphasis of this meeting was on future directions, it is not surprising that there are many obstaclesto full realization of the potential of most of the areas of research discussed in these reports. In many, such asthe matter-field states in which external field and binding field strengths are comparable, there are serioustheoretical and computational difficulties that must be overcome in order to achieve understanding andpredictability. In others, the experimental techniques and instrumentation are just now reaching a level ofsophistication that enables “proof of principle” experiments, and considerable development will be requiredbefore real investigations and exploitation of the field can occur.

Overall, the picture that emerged from this workshop was of fields that are extremely dynamic and exciting,and critical to the development of many other fields of science. Important new phenomena, new theories,new techniques, new understanding were to be found in all of the areas that were discussed. Development ismoving at an unprecedented rate in many of these areas. The challenges, the opportunities, and theexcitement were clearly described by the participants, and well captured by the contributions contained inthis volume.

Lloyd Armstrong, Jr., Chair

Atomic, Molecular, and Optical Physics Workshop Report

7 Panel A

Panel A: Interaction of Atoms and Moleculeswith Photons - Low Field

Chair:Daniel Kleppner (Massachusetts Institute of Technology)

Panelists:Uwe Becker (Fritz Haber Institut der MPG)Nora Berrah (Western Michigan University)

William Cooke (College of William and Mary)Robert Field (Massachusetts Institute of Technology)

Wallace Glab (Texas Tech University)Wendell Hill (University of Maryland)

Eric Horsdal-Pedersen (Aarhus University)Stephen Leone (University of Colorado, JILA)

Marsha Lester (University of Pennsylvania)Stephen Lundeen (Colorado State University)

Cheuk-Yiu Ng (Iowa State University)Erwin Poliakoff (Louisiana State University)Stephen Pratt (Argonne National Laboratory)William Stwalley (University of Connecticut)

Michael White (Brookhaven National Laboratory)

Outline Summary

1. Atomic and Molecular Manipulation and Control

2. Advanced Spectroscopies with New Light Sources

3. Quantum Dynamics of Interacting Systems


Panel A 8

Summary

Laser cooling and trapping techniques, synchrotron lightsources, advanced imaging methods and otherinnovations have opened frontiers for research in atomicand molecular physics. Opportunities encompassfundamental studies of structure and dynamics, thedevelopment of knowledge and data for other disciplinesand for industry, and the development of newtechnologies. We describe opportunities in threecategories.

1. Atomic and Molecular Manipulation andControl

Light force techniques make it possible to coolatoms to sub-microkelvin temperatures and tostudy ultra cold collisions and new types ofmolecular association. They have led to thecreation of matter-wave optics and to Bose-Einstein condensation in an atomic gas. Thesedevelopments have created opportunities for thestudy of Bose-Einstein condensates, the creationof new condensate systems, the development ofatom lasers, and new studies of fundamentalsymmetries. The cooling methods may beextended to molecular systems to create new typesof spectroscopy and studies of dynamicalprocesses. Applications for the research exist inareas ranging from biology to nanotechnology. Inaddition, new techniques to control andmanipulate ions, including ion storage rings, canbe employed to study molecular ionic systems thatare important in the atmosphere and in low energyplasmas.

2. Advanced Spectroscopies with New LightSources

Two developments have stimulated the creation ofnew spectroscopies and new methods for thequantum control of atomic and molecularprocesses. These are the creation of thirdgeneration light sources and practical methods forgenerating ultrashort optical pulses. Researchopportunities with third generation light sources

include studies of the basic physics of photon-matter interactions, new investigations of themany-body problem, studies of the productdistributions from photo-initiated dissociation, andstudies of the structure and dynamics of highlycharged ions. The studies are germane toproblems ranging from the origin of radiationdamage in cells to hot plasmas in stars. Ultrashortoptical pulses can be tailored to study atomic andmolecular wavepacket dynamics. Tailored pulsescan be used to control molecular dynamics and theproducts of chemical reactions, and can providenew diagnostics for identifying atoms andmolecules in hostile environments.

3. Quantum Dynamics of Interacting Systems

Understanding chemical reactivity requiresunderstanding the quantum dynamics ofinteracting systems. Scattering, including photoscattering and photo ejection, is the chiefexperimental tool for studying quantum dynamics.The development of novel photon sources andtime- and position-sensitive imaging and co-incidence methods opens the way to study themechanisms of energy exchange within amolecule. Reactions between simple systems canbe observed with a detail approaching completequantum specification. There are opportunities forextending these experiments to more complexsystems and for developing visualizationtechniques for interpreting the voluminous data.In addition to elucidating the basic processes ofchemical reactivity, such studies are important toastrophysics, plasma physics and materialsscience.


9 Panel A

1. ATOMIC AND MOLECULARMANIPULATION AND CONTROL

1A. Introduction

The underlying goal for much of the research inatomic, molecular and optical physics (AMOP) isto understand atomic, molecular and opticalprocesses with the ultimate clarity that quantummechanics allows. Achieving this goalexperimentally requires tools for preparing atomicand molecular samples in specific states, and forcontrolling their motion. Recently there has beena revolution in techniques for controlling not onlythe internal states of atoms but also theirtranslational states, providing qualitatively newlevels of control. These methods have openedfrontiers of research, generated strong links toother areas of science, provided new opportunitiesfor metrology, and are starting to have an impacton technology.

1B. Light Forces—Manipulating Matter byLight

Motion is an impediment to the study of atomic,molecular and ionic structure and dynamics. Motionblurs spectral lines and obscures structure. It limitsobservation time, degrades the control of collisionkinetics and introduces undesirable complexitiesinto observations. For this reason, experimentalprogress in AMOP has often been propelled byprogress in controlling atomic and molecularmotion. Early efforts to control motion employedeffusive atomic beams and mechanical velocityselectors. Supersonic beams were developed thatfurther reduced thermal effects in atoms andmolecules, creating a new arena for studying atomicand molecular dynamics and also the science ofclusters, including the discovery of the fullerenes—honored by the 1996 Nobel Prize in chemistry. In aparallel development, the creation of ion traps led toessentially total control of ionic motion and openedthe way to advances in spectroscopy, collisiondynamics, plasma processes, and metrology.

Within the past decade a revolution in atomiccontrol has been precipitated by the developmentof light force techniques. The significance of thisrevolution was dramatized by the award of the1997 Nobel Prize in physics for the developmentof atomic light force methods. The essence of theadvance is this: In the past, translational motionhad to be treated classically, and kinetic energywas unquantized. Today, the ultimate limit intranslational control has been achieved; it is nowpossible to put atoms into a single translationalquantum state and to observe their behavior incomplete detail.

The new methods employ laser techniques forslowing, cooling, and manipulating atoms, andoptical and magnetic traps for confining them. Thetraps can hold large numbers of atoms at sub-microkelvin temperatures. These methods make itpossible to witness the photo-association of freeatoms into unusual molecular states; to studyultracold collisions in which the dynamics can becontrolled by electric, magnetic, and laser fields;to observe phenomena such as Blochoscillations—collective oscillations in a periodicpotential—and to realize the quantum kickedrotor—a paradigm in the study of nonlineardynamics.

The observation of Bose-Einstein condensation(BEC) in an atomic gas is the most dramaticachievement of the light-force techniques. Theclarity with which the Bose condensates can beobserved and the exquisite control by which theycan be manipulated, as well as the novelty of thephenomena, have generated a new field of studythat has attracted enormous interest in AMOP bycondensed matter scientists, chemists, and others.The phase transition, the collective motions andtransport and relaxation processes of thecondensate, have all been studied. A study of thephase coherence of the condensate resulted in thecreation of an atom laser.


Panel A 10

Closely related to these developments is thecreation of the field of matter wave optics.Several types of atom-wave interferometers havebeen demonstrated and optical elements such asmirrors, lenses, gratings, beam splitters, andhollow guiding fibers, have been developed. Theinterferometers have been employed to measureatomic and molecular properties, to investigate thedecoherence in quantum systems, to measurefundamental constants, and to create ultrasensitiveaccelerometers. Atom-optics methods are beingapplied to new types of lithography, and havealready been used to create length standards forscanning microscopies.

Scientific opportunities and applications include:

• Addressing fundamental questions on BECincluding transport, condensate dynamics,coherent collisions, compound condensates,and interactions with light. The condensatesmay make possible new studies in quantumentanglement.

• Exploring new condensates and developingmethods for increasing the size of BECcondensates. Such methods could beapplied to much more intense atom lasers.

• Developing the atom laser as a scientifictool, and exploring its applications to nano-technology.

• Extending photoassociative spectroscopy ofsubmillikelvin atoms to study long rangemolecular states of other atoms (for instancethe hyperfine-free alkaline earths, andmetastable rare gases) and also to studyatom-molecule and molecule-moleculephotoassociation. Such studies provideparameters such as scattering lengths,dissociation energies, and spin-flip crosssections. These are critical forunderstanding very low temperaturedynamics and BEC.

• Developing light force and trappingtechniques for molecules. Trapped moleculemethods would provide new ways to studymolecular structure, chemical reactions, andinteratomic and inter- molecular interactionsthat may ultimately yield BEC of state-selected molecules, and molecule lasers.

• Applying low temperature molecularspectroscopy to new studies in molecularand biological physics. For example, lowfrequency torsional modes in biopolymerscan be cooled to ultralow temperatures,perhaps yielding new insights into proteinfolding.

• Developing light force techniques for otherapplications. A notable example is the earlydevelopment of optical tweezers, which areused in polymer studies, biology, and otherapplications. These tweezers are able for thefirst time to directly manipulate matterwithin a living cell and to stretch and evenbreak an individual polymer or DNAmolecule.

1C. Fundamental Studies

Closely related in spirit to the study of light forcesis a body of research in AMOP centered on thestudy of fundamental symmetries and otherfundamental phenomena. The goal is to deepenour basic understanding of physics, but inaccomplishing this the research has often pushedforward the frontiers of experimental control, andof metrology.

Among the most precise studies in AMOP are aseries of experiments on paritynonconservation (PNC) effects in atoms. Thesehave contributed new information to the StandardModel and they have led to the discovery of a newtype of nuclear current—the anapole moment.Laser power buildup cavities for these experimentshave advanced laser mirror technology indevelopments that led to cavity ring-down


11 Panel A

spectroscopy and that have been taken up byindustry. New opportunities for this line of researchhave been opened by advances in light forcetechniques and in the trapping of radioactive atoms.

Atomic studies of the fundamental symmetriesprobe the basic forces that govern the interactionsin nature and search for new types of fundamentalparticles. Such investigations range from theextremely high precision spectroscopy of neutraland ionic atoms and molecules to the creation ofexotic atoms and antimatter. These experimentsgenerally push the frontiers of experimentaltechniques and are a source of new technology forother sciences and for applications. Studies ofQED in atoms are spurring methods to measureoptical frequencies that have potential applicationsin communication and control. Searches forelectric dipole moments of fundamental particleshave led to the development of magnetometers andimproved methods for measuring and controllingelectric and magnetic fields. Investigations ofantimatter atoms have been instrumental in thedevelopment of ion trapping.

In a different domain, new types of light have beencreated with unique quantum properties. Theseinclude squeezed light, correlated photon beams,and quantum solitons. A common feature is thatthe fluctuations lie below the so-called standardquantum limit associated with the vacuum state ofthe electromagnetic field. In addition to theirfundamental interest, such nonclassical lightsources may have spectroscopic applications.Nonclassical light has enhanced spectroscopicsensitivity beyond the standard quantum limit. Ifthis technique can be made practical, it offersopportunities for advances in ultrasensitivedetection and it may increase the sensitivity ofgravity wave detectors such as LIGO. Closelyrelated to these developments are the creation ofnew types of photon and atomic states using themethods of cavity quantum electrodynamics.

Opportunities include:

• Exploiting atom trapping and coolingtechniques for tests of PNC interactions andother fundamental tests.

• Developing new methods for thespectroscopy of short lived atomic species.

• Applying optical methods to polarize atomicnuclei to create polarized beams and targetsfor nuclear and particle physics. Recentstudies on the spin structure of the neutronthrough electron scattering from polarizedhelium-3 nuclei provide a notable example.

• Finding more efficient methods forpolarizing rare gas nuclei. The productionof these nuclei have led to the developmentof an important new medical diagnostictechnology—magnetic resonance imagingwith polarized rare gases—with greatlyenhanced sensitivity and resolution.

• Investigating the behavior of atomic systemsin nonclassical light fields to understandaltered lineshapes and decay rates, nonlinearmixing, and new types of lasing behavior.

• Studying entangled atomic states using themethods of cavity quantum electrodynamicsto obtain new techniques with potentialapplications in the general area of quantumcomputing and quantum cryptography.

1D. Molecular Control

Molecular spectra can be vastly simplified, andnew types of molecular states accessed bymethods such as multiple optical-optical andoptical-RF resonance techniques, polarizationlabeling spectroscopy, multiphoton ionization,laser-induced fluorescence, and stimulated Ramanadiabatic rapid passage. Beyond these, molecularspectra can be greatly simplified by cooling themolecules. Spectroscopy has been carried out totemperatures as low as 1K using seeded


Panel A 12

supersonic molecular beams or solid rare gasmatrices. More recently, temperatures as low as0.4K have been achieved, for example by studyingspecies in helium-4 clusters.


• Extending methods for manipulating andcontrolling the vibrational populations ofmolecules to their rotational and electronicstates. Such control would provide newpathways to the studies of collisiondynamics and chemical reactions.

• Developing methods for cooling andtrapping molecules. These methods couldprovide new levels of resolution and vastlysimplify molecular spectroscopy andprovide a new arena for the study ofcollisions and reactions.

• Controlling detrimental loss processes intraps by controlling collision dynamics.

1E. The Control of Ions

Ions can be manipulated and controlled with trapsfor collision studies, ultrahigh precisionspectroscopy, studies of ion crystals and stronglycoupled plasmas, and for applications to quantumlogic gates and quantum computing. Ion trapmethods have also been used to observevibrational relaxation of molecular ions, anddissociative and dielectronic recombination at verylow energies. Because dissociative recombinationcan play an essential role in determining thedistributions of free radicals in low energyplasmas, this research can have importantapplications in materials processing and etching.

In heavy ion storage rings molecular ions can bestored at MeV energies for several seconds,allowing them to cool rotationally andvibrationally. Reaction products from dissociativerecombination and branching ratios for electronicstates can be identified using imaging techniques.

Ions that can be conveniently studied include O2+

and N2+, which are relevant to the physics andchemistry of the Earth’s atmospheres (N2+ is alsoimportant on Mars), and H2+, which is importantto the chemistry of the interstellar space.


• Producing state-selected molecular ionsdirectly from ultracold atoms byphotoassociative ionization.

• Manipulating ions in storage rings by laserlight, providing a new control over reactiondynamics.

2. ADVANCED SPECTROSCOPIESWITH NEW LIGHT SOURCES

Two developments are stimulating the creation ofnew forms of spectroscopy and new methods forthe quantum control of atomic and molecularprocesses. These are the creation of thirdgeneration light sources to extend the power andresolution of spectroscopy in the UV and short x-ray regimes, and the development of practicalmethods for generating tailored ultrashort opticalpulses. Additional opportunities due to freeelectron laser light sources are discussed by thePanel on Interactions of Atoms and Moleculeswith Photons: High Energy.

2A. Third Generation Synchrotron Sources

The photon beams now available from thirdgeneration synchrotron light sources developed atBerkeley (ALS) and Argonne (APS) offer newopportunities to study radiation-matter interactionsthat require very high energies, spatial resolution,spectral resolution, or polarization. The lowemittance of third generation storage rings makesit possible to employ undulator or wiggler devicesthat can provide radiation into the hard x-rayregion with unprecedented intensities and spatialcollimation. In atomic and molecular physics,


13 Panel A

these developments open new avenues forstudying multi-electron excitation and inner-shellphotoionization. For example, high brightnessundulators in the UV can be used for “universal”detection of neutral product species via one-photon ionization with sensitivities rivalingconventional electron-impact sources, but withoutthe associated difficulties of unwanted dissociativeionization and a hot source region.

Complex atomic, molecular and ionic processes canbe studied by combining third generationsynchrotron radiation with high resolution detectiontechniques that simultaneously detect and analyzeevery possible outcome channel. These methods alsomake it possible to image the fragments emitted inphotoprocesses-ions, electrons, neutral, andphotons—and to detect their coincidences. Becausepolyatomic molecular species can exit in differentchemical isomers with distinctly different chemicalproperties and energetics, the detailed understandingof photodissociation, electron dissociation, andreaction processes is possible only with such anidentification of nascent product chemical structures.


• Testing the basic assumptions that have beenused traditionally to interpret photon-matterinteractions; for instance, the independentparticle, impulse, and dipoleapproximations.

• Conducting quantum mechanically completeexperiments that determine the alignmentand orientation parameters for atomic andmolecular targets, using circularly orelliptically polarized light from helicalundulators.

• Advancing our understanding of the many-body problem, one of the central problemsin atomic, nuclear and condensed matterphysics, through negative ion spectroscopy,as suggested by theoretical studies.

• Identifying nascent product distributions ofatomic and molecular fragments, resultingfrom reactive scattering, electronattachment, and photo-initiated dissociation,ionization, detachment and desorption.

• Studying endothermic reactions with low crosssections and other processes that result indifficult to detect, low-yield neutral products.

• Combining ultra-high resolution capabilitiesof small aperture monochromators with thebeams from undulators to produce narrow-band UV and soft x-ray radiation to studyphotoionization of atoms, molecules, clustersand ions, possibly with complete quantumstate specification and kinematic imaging.

• Employing high resolution photoelectron-photoion coincidence techniques, to prepareatomic and molecular ions with specifiedquantum states for scattering andspectroscopic experiments. Neutral atoms(O, N, and C, etc.) and radicals (OH, CN,CH3, and C2H, etc.) can be prepared byphotodissociation of appropriate precursormolecules using the high flux VUVsynchrotron radiation.

• Employing light from third generationsynchrotron sources to create new species;for example, “hollow” atoms in which thecore electrons are simultaneously excited tohigh-lying states. Such atoms provide a newarena for observing many-electron systemsand testing theories of highly correlatedelectron motion.

• Studying the structure and dynamics ofhighly charged ions by combining radiationfrom third generation synchrotron sourceswith targets of highly charged ions. Thesestudies will provide precision data on ionicstructure and electron-correlation effects,benchmarks for theoretical studies ofmultielectron interactions and data forplasma modeling and ion-beam lithography.


Panel A 14

2A-2. Related Science and Applications

X-ray absorption spectroscopy with third-generation synchrotrons can provide element-specific and chemical state-specific detection oftrace contaminants in biological materials, soilsand groundwater. The spectroscopy can beapplied to understand radiation damage incomplex molecules such as DNA, for the damageoriginates in core excitations that arefundamentally atomic processes. A realunderstanding of radiation damage is possible onlythrough studies of the complex decay processes inisolated atoms and simple molecules.

Studies in the structure and dynamics of highlycharged ions is vital for understanding hot plasmasin stars, in fusion reactors, and also in x-ray lasers.

The energetic, spectroscopic and dynamicalinformation from molecular studies with thirdgeneration synchrotron sources can be importantto combustion chemistry, plasma science, andenvironmental and atmospheric chemistry. Forinstance, cation spectroscopic studies can yieldinvaluable spectroscopic and geometricinformation on the neutral ground state.Rotationally resolved photoionization andphotoelectron spectroscopic studies of reactiveintermediates, such as CH, CH2, NH, NH2, HCO,CH3O, and C2H, are important in combustion,plasmas, and atmospheric chemistry.

2B. Ultrafast Lasers and Wave PacketDynamics

Sub-picosecond laser systems have becomepractical laboratory instruments. Theirdevelopment has generated an explosion of newresearch on the control of elementaryphotochemical and photophysical processes byusing sequences of ultrafast pulses with preciselyshaped spectral and temporal profiles. The pulses,which are shorter than the period of typicalmolecular vibrations and rotations, or the time

between collisions in a dense gas, make possiblenew ways for manipulating and interrogating theinternal dynamics of molecules. Several groupshave employed shaped optical pulses to generateatomic and molecular wave-packets. Thesetechniques may be able to provide such precisecontrol over internal interactions that the finalproducts of chemical reactions can be selected bysuitably arranging a sequence of shaped pulses. Inrelated control experiments, interference betweentwo optical excitation pathways has been used tocontrol the branching ratios betweenphotoionization and photodissociation in hydrogeniodide, and interference between one- and two-photon excitations has been used to control thephotoionization angular distribution and thedirection of the flow of current in solid statedevices.

In a somewhat different context, new types ofwave-packets have been created in Rydberg atomsby employing combinations of laser light,microwave fields and time-varying electric andmagnetic fields. The electronic current can bestationary and highly localized, as in a “circular”Rydberg state, and the electron can even belocalized in a stationary wave packet, as in acoherent elliptic state. Such atomic states havealready been employed in studies of quantumoptics, high resolution spectroscopy, andscattering. They have also been employed to studythe physics of the semiclassical regime, a problemwhich is of deep interest in molecular physicswhere semiclassical approximations are oftenappropriate.


• Extending these tailored pulse techniques bycombining several shaped, multicolorpulses.

• Investigating wave packet dynamics byexperimentally executing algorithms createdwith optimal control theory.


15 Panel A

• Employing short-pulsed, multi-stepexcitation to create a new collision-free toolfor identifying atoms or molecules, even inhostile environments such as plasmas orflames.

• Investigating the possibility of preparingspecies in wave packets to control thedynamics of interactions, i.e., of controllingmolecular reactions.

• Developing new methods for creating andstudying the physics of localized atomicstates, and for applying these states toscattering problems and studies ofsemiclassical physics.

3. QUANTUM DYNAMICS OFINTERACTING SYSTEMS

Understanding the fundamentals of chemicalreactivity demands understanding the quantumdynamics of interacting systems. Such anunderstanding would contribute not only tofundamental chemical dynamics but to essentiallyevery area of chemistry and related disciplinessuch as atmospheric science, combustion,astrophysics, plasmas, and materials science. Thequantum dynamics of interacting systems holdsthe key to controlling and manipulating reactions.It is difficult to overestimate the power that suchcontrol would provide.

Just as spectroscopy is the chief experimental toolfor studying the structure of atomic and molecularsystems, scattering—including photon scatteringand half-scattering (photo ejection processes)—isthe primary experimental tool for studyingdynamics at the quantum level. Spectroscopy isalso an important tool for studying dynamicalprocesses such as photoassociation andphotoionization, and it can provide essentialdiagnostics for scattering experiments. And just asspectroscopy has undergone a revolution, so toohas scattering, for the clarity with which scattering

events can be observed, the variety of systems thathave become accessible, and the range indynamical variables with which they can bestudied, have expanded enormously.

3A. Complete Scattering Experiments

As its name suggests, a complete scatteringexperiment provides all the information about ascattering event that quantum dynamics allows us toknow. Instead of an average transition rate, one talksof the quantum mechanical amplitude for passingfrom one quantum state of the collision partners toanother. Such an experiment requires precisemeasurement and control of the incoming andoutgoing momenta of the scattering partners, and alsotheir polarization, alignment, and orientation. Withsuch measurements one can obtain both themagnitudes and phases of the scattering amplitudesfor targets and projectiles such as atoms, ions,photons, electrons, and simple molecules.

Although a complete scattering experimentremains an ideal, even in cases where it cannot befully achieved the concept can nevertheless beemployed to distinguish the important degrees offreedom in complex systems. Techniquesemployed in complete scattering experimentsinclude polarized beams and targets, spin resolvedelectron detection, and laser-selected molecularstate preparation and detection. The research hasbeen propelled by the recent development of time-and position-sensitive imaging methods thatprovide what are known as 4 π detection, detectorsin which every scattering partner in a collision isobserved and fully measured. Their highcollection efficiency can make coincidenceexperiments feasible. These detectors have beenapplied to photo fragmentation (“half collisions”)and particle-induced reactions involving atomsand simple molecules, for instance, to thephotodissociation of molecular hydrogen. Otheradvances include photofragment imaging andmultiphoton ionization of neutral products.


Panel A 16

Scientific opportunities and applications include:

• Extending complete scattering studies tomore complex systems such as polyatomicmolecules, radicals and ions. Theseexperiments, with their full specification ofinitial and final quantum states, can revealthe mechanisms for energy exchangebetween electronic, vibrational, rotationaland translational degrees of freedom. Suchexperiments have been carried out onreactions between atoms and simplediatomic molecules, for instance deuteriumand molecular hydrogen, and on thecompetition between autoionization andpredissociation in highly excited smallmolecules.

• Extending these techniques to clusters.Collision experiments with completespecificity are not yet possible due to thehigh density of final quantum states in acluster, but much can be learned fromcollisions with electrons, atoms, and ions, inexperiments employing mass spectrometryand time-of-flight energy analysis. Suchexperiments can reveal, for instance, themass distribution of photofragments,information that is essential forunderstanding cluster-cluster and cluster-surface interactions.

3B. Visualization of Quantum States andDynamics

A serious problem in interpreting data fromcomplete scattering experiments and comparingthe results with theory is that even relativelysimple systems require complex multidimensionalrepresentations. It is essential to find methods toview this information in ways that yield insightsinto the essential degrees of freedom, time scales,and the various competing mechanisms. Theparticular way in which the results are presented iscrucial in determining whether or not the essentialprocesses can even be recognized.

In some cases, a subset of the coordinates can beemployed in a reduced dimension visualizationscheme. An example is the femto-chemists’picture of a wave packet moving on a potentialsurface. This sort of picture aids in the design ofexperiments and in the identification of physicallyplausible coupling mechanisms. Non-separable,nonperturbative systems are often separable for ashort time after a short pulse preparation of alocalized excitation. In some cases the first stagesof an interaction can be separated into active andpassive degrees of freedom. In such cases thereduced dimension representation of the evolvingsystem can show the evolution within a system ofa well understood subsystem that has beenpreviously studied in isolation.

In high resolution frequency domain spectroscopy,a spectrum can consist of entangled patterns arisingfrom different values of a conserved observable.By comparing different spectra in which theentangled patterns appear with different relativeweights, the dynamically independent pieces of thespectrum may be separated. For example, data onthreshold double ionization of helium presented interms of the momentum vector of the relativemotion of the two electrons revealed a simpledynamical pattern that was essentially invisiblewhen presented in terms of the conventionallaboratory frame momentum vectors.

Scientific opportunities include:

• Developing reduced-dimension visualizationmethods for optimally representing acomplex system in terms of weaklyinteracting subsystems. It is essential to findwhen such representation are appropriate,for they can introduce major simplifications.

• Developing pattern recognition techniquesthat can recognize an unspecified andunknown patterns in a large, high qualitydata set.


17 Panel A

• Developing methods that reveal correlationsin the internal relative momenta in multi-dimensional velocity space. Such methodsare needed for understanding experiments inwhich the momentum vectors of all finalreaction products are determinedindividually event by event.

• Investigating the use of collective, Jacobi, orother appropriate internal momentumcoordinates, to reveal patterns that reveal theroles of two-body or collective interactions.

3C. Cross-Disciplinary Connections

Advancing our understanding of the quantummechanics of interacting systems has broadramifications and the impact of the research is noteasily summarized. Nevertheless, it is of value tonote some of these.

a. Comprehensive sets of dynamical parameters ofatomic collision and half-collision processesprovide information needed to derive partial cross-sections, angular distributions, and spinparameters of a large variety of atoms andmolecules that are important in astrophysics,plasma physics, and materials science.

b. The interaction of highly charged atomic ionswith atomic, molecular, electronic and ionictargets occupies a special place in the study ofquantum dynamics of interacting systems becauseof their importance in high temperature plasmas.Such plasmas occur in such fusion plasma devicesand laser-produced plasmas in both ICF and x-raylaser work. Opacities, charge-state equilibria andequations of state of such plasmas, as well asenergy flow rates through short lived plasmas, canbe calculated if relevant recombination, electrontransfer, and ionization rates are known.Interactions involving ions in states appropriate toplasmas with electronic temperatures of a fewhundred eV to over 100 keV are particularly

important. In addition, a quantitativeunderstanding of binary ion-ion and photon-ioninteractions is needed for numerous plasmadiagnostic techniques.

c. Understanding theoretically and experimentallythe transition of scattering characteristics as afunction of collision energy where the collisionevolves from molecular characteristics (e.g.,molecular wave functions, potential curves, etc.)to that of proceeding via atomic characteristics(e.g., impulsive interactions, atomic states, etc.) isimportant to basic physics and for applications.

Panel B: Interactions of Atoms and Moleculeswith Photon - High Field

Chair:Phil Bucksbaum (University of Michigan)

Panelists:Louis Dimauro (Brookhaven National Laboratory)

Richard Freeman (Lawrence Livermore National Laboratory)Anthony Johnson (New Jersey Institute of Technology)

Robert Jones (University of Virginia)Jeffrey Kimble (California Institute of Technology)

Wolfgang Sandner (Max Born Institute/Berlin)Carlos Stroud (University of Rochester)


19 Panel B

I. Executive Summary

In “high-field” interactions applied field energies equalor exceed the energy density of matter, so that the fieldcontrols the shape and dynamics of atoms, molecules,or other material systems. Using modern tools of high-field and ultrafast physics, one no longer merelyobserves nature, but can reshape and redirect atoms,molecules, particles, or radiation. This new drivetowards harnessing quantum dynamics is enormouslyimportant to future developments in fundamentalphysics and applied energy science.

This document explores some of the ramifications andfuture fertile directions in this field. We divide thepaper into three “grand visions,” which are broadinterconnected areas of AMO science. We begin withan executive summary:

I.A. Frontiers of Time, Intensity, andWavelength

The first vision is the technical frontier of high fieldsand ultrafast optical pulses. The lasers that producethe intense fields are themselves designed and builtusing the principles of this new science ofmanipulating matter.

The past decade has seen a thousand-foldimprovement in the peak power of solid-state lasers,and the future holds challenges for scaling to higherpower with both solid-state and proposed free-electronsources. This progress will enable a wide range ofscience: new methods for more efficient inertialconfinement fusion; compact laser-driven electronaccelerators; sources of femtosecond (10-15 seconds)gamma radiation are some of the prospects.

Pulses will also become shorter, opening new areas ofscientific inquiry previously undreamed. Femtosecondpulses, which were frontier research projects ten yearsago, have been commercialized. Attosecond (10-18

seconds) pulses are actively investigated, and theircontrol and measurement is a major challenge.Attosecond science will certainly center on the

dynamics of atoms and molecules, since the time forscattering between a relativistic particle and an atom ison this timescale.

High-field laser-atom interactions have recentlyyielded intense coherent radiation at wavelengths asshort as 27 Angstroms, which is short enough forbiological imaging of femtosecond dynamics in the“water window.” These light sources are majorcomplements to synchrotron sources, deliveringshorter pulses and greater control over coherence, onplatforms that can be as small as a single table top.Still shorter wavelength sources have beendemonstrated by scattering high intensity laser pulsesfrom relativistic electon beams, or gating synchrotronradiation using laser-driven material switches. Theseadvances bring kilovolt, megavolt, or even gigavoltphotons into reach for laser-based time-resolvedscience of materials, chemistry, atomic physics, andpossibly even particle physics.

We now control not only the wavelength, intensity, andduration of intense fields, but also their detailed shape,often down to the single-cycle limit. Pulse shaping isa major enabling element of the second “grand vision,”which we call:

I.B. Harnessing Quantum Dynamics

Quantum dynamics in an atom or a molecular bondinvolves atomic-size fields (volts/angstrom), times(picoseconds and below), and frequencies (terahertz[1012 hertz] to exahertz [1018 hertz]). Our challengeis to achieve the technical ability to manipulate thesestates, the knowledge of how to design new“sculpted” states, and the wisdom of which new stateswill best accomplish our goals of new insight anduseful applications.

Some of the frontier areas are fundamental, such asnonclassical properties of entangled states of the fieldsand matter, or the control of chaotic behavior innonseparable dynamical systems. Others challengeslie at the interface between basic knowledge andtechnology. For example, adaptive feedback could


Panel B 20

permit the newly formed quantum state to “teach” theexternal field how to alter its own shape. Sometimesthis adaptive feedback is built into the quantumsystem, such as in relativistic plasma channeling orelectromagnetically-induced transparency. An evenmore exciting vision for the future, however, is usingfast computers and pulse shapers to alter intense fieldsaccording to instructions given by the quantum state,in order to optimize the state for a particular purpose.Some of the possible challenging applications of field-sculpted matter are:

• Bond-selective chemistry, in which a strongprogrammed optical field creates a rovibrationalor electronic wavepacket in a molecule, whichinduces a specific change in unimolecularchemistry;

• Coherent collisions, where target or projectileare prepared in wavepacket states to controlspecific inelastic channels or select impactparameters;

• Quantum encoding, the encrypting ofinformation into the amplitude and phase ofnonstationary states. A strong programmed fieldpulse can key the return of the information as aquantum echo;

• Altering bulk properties, to control Braggscattering, switch crystal domains, or movecharge in solids and plasmas;

• Energy channeling, the controlled transport anddeposition of laser energy into media which aregenerally opaque to weak or unshaped opticalpulses. The mechanisms for energy channelingchange dramatically when the intensity is veryhigh, and this leads us to the third grand vision:

I.C. Matter in extreme conditions

At the extreme limit of high intensity in a laser focus,the electric fields can exceed trillions of volts percentimeter; magnetic fields can approach gigagauss;pressures aproach a gigabar. Physics in thisenvironment is largely unexplored.

The high intensity can be converted to high pressure insolid-density targets, to produce conditions usuallyonly observed in astrophysical objects or inthermonuclear reactions. Possible applications areadvanced x-ray sources, or creating initial conditionsfor inertial confinement fusion.

New nonlinear effects occur when the work done onan electron during one optical cycle exceeds theelectron rest mass (intensities exceeding one exawattper square centimeter). Optical propagation inplasmas becomes dominated by the oscillatingrelativistic mass of the electrons. This coherentmotion creates extreme charge-density gradients inplasmas, which can support fields of gigavolts percentimeter. Nonlinear Compton scattering in thisregime creates harmonics from the relativistic motionof the electrons. Photoionization into such arelativistic continuum has not been studied.

At still higher fields, near the material limits for focusedlaser light in the laboratory but well beyond what hasbeen demonstrated, electrons can be driven so violentlythat they emit electron-positron pairs through theprocess of coherent nonlinear Compton scattering.

Still higher intensities can be achieved in the center ofmomentum frame by back-scattering laser light from arelativistic electron beam, such as the SLAC linac. Inthe CM frame of the backscattered gamma ray and acounter-propagating laser beam, the field can rise toabove ten million gigavolts per centimeter, at whichfield the vacuum itself becomes unstable, decaying intoan electron-positron plasma. Positrons from thismechanism were recently observed at SLAC (D.L.Burke, et al., PRL 79, 1626 (1997)). This was the firstexample of coherent sparking of the vacuum.

I.D. Reading List and References.

This document was prepared at an intense two-dayworkshop in Chantilly, VA. As such, it can onlyprovide qualitative and general information on thesubject. There are, however, several importantfocused conferences in this field where current work


21 Panel B

and future prospects are discussed in detail, and thereader may find additional details and documentationin their proceedings:

• Super-intense laser-atom physics (SILAP), mostcurrent meeting was in Russia in August, 1996.Proceedings edited by H.G. Muller and M.Federov.

• OSA Topical Meeting on Short Wavelengths andHigh Fields, most current meeting in Santa Fe inMarch, 1997, Proceedings edited by L.DiMauro and M. Murnane, not yet published.The previous meeting in this series is available:“High Field Interactions and Short WavelengthGeneration, AUG 1994, St. Malo, France,Paperback, ISBN 1-55752-361-4, Availablefrom the Optical Society of America.

• International Conference on MultiphotonPhysics (ICOMP), most recent meeting inMunich in October, 1996, Proceedings edited byP. Lambropoulos.

• 10th ICFA Beam Dynamics Panel Workshop;4th Generation Light Sources, Grenoble, FranceJan 22-25, 1996, editor J.L. Laclare. The reportis available on the web: http://www.esrf.fr/conferences/conclusions/webconc.htm

• Proceedings of the Seventh AdvancedAccelerator Concepts Workshop, Lake Tahoe,CA, Oct., 1996, W. Chattopadhyay, J.McCullough and P. Dahl, Editors, AIP,conference Proceedings 398 (American Instituteof Physics, New York, 1997).

• OSA Topical Meeting on Ultrafast Processes,ULTRAFAST PHENOMENA, MAY 1996, SanDiego, CA, ISBN 1-55752-441-6, availablefrom the Optical Society of America.

• University of California at Berkeley Fast IgnitorPhysics Workshop, March 1997. Unpublishedproceedings might be available from Dick Lee atLawrence Livermore National Lab.

II. Opportunities in Intense-Field AMOPhysics

Under high-field conditions, systems are driven byexternal fields to evolve more rapidly than theirnatural dynamical time scale. In quantum mechanicallanguage appropriate to atomic and molecular physics,we state this as follows: A dynamical system u(t)evolves according to a Hamiltonian

(H + H’(t))u(t) = i hbar [d(u(t))]/dt

Where H is the Hamiltonian of the system excludingthe field interaction, and H’ is the interactionhamiltonian. When H’ evolves on a time scale fasterthan the natural dynamical timescale of u(t), and/or hasa magnitude comparable to or greater than H, then weare in the high field regime. Typically, traditionalapproaches to dynamics in quantum systems such asperturbation theory do not reveal the physics of suchsystems.

Instead, the field and the matter form together a newsystem which generally has altogether new properties.Some of these properties will be critical for advancesin energy science and technology, and these areas formthe grand challenges of this field.

II. A. Fr ontiers of Time, Intensity, and Wavelength

In this section we will discuss the physical frontiers inmaterials, concepts, and optics at high fields, shortwavelengths, and short pulsewidths. The generation ofultrahigh fields continues to go hand in hand with thegeneration of ultrashort pulses. The current state-of-the-art in solid state lasers operating at wavelengthsnear 1000-nm have yielded pulses as short as 5 fs andintensities 1022 W/cm2. Currently, the intensity islimited primarily by the saturation intensity andgeometry of the gain medium. The pulse duration isclose to the limit of the uncertainty principle with a 5fs pulse at 1000 nm consisting of approximately 3optical cycles. If future technology could scale thecurrent 1 cm diameter solid state lasers to 1 m, thenintensities of 1026 W/cm2 and concomitant fieldstrengths of 1014 V/cm should be achievable.


Panel B 22

If these technological extremes (e.g., scaling the sizeof gain medium) are achieved, where will the nextadvances come from? Laboratory intensities evenabove the solid state laser intensity limit of 1026 W/cm2 might be reached by future Free-Electron Lasersources. Alternatively, the problem of scaling can besidestepped altogether by the use of relativistic targets,which see the optical fields upshifted by a factor ofgamma. Although many processes cannot be studiedthis way, relativistic electron and ion beams havealready been used to study special regimes inhigh-intensity physics.

In the area of ultrashort pulse generation, if we pushthe limits of high harmonic generation for example tothe 1000th harmonic of a solid state laser to 10-Angstroms (~ 1 keV) and assumed we could phasematch this process then we would predict pulses oforder 10 as (attoseconds). We have always inferredthat the timescale for electron orbits was on this order,but we may soon have the capability to directlymeasure it and predict new phenomena!

High intensity ultrashort pulses are necessary togenerate coherent sources of x-rays. Recent work bytwo groups have demonstrated coherent high harmonicgeneration within the water transparency window ofcritical importance for imaging biological processes.Among the many uses for this unique source of lightwould be the following: x-ray holography for medicalimaging applications; table-top sources couldcomplement synchrotrons as the sources of lightnecessary for x-ray lithography to push the limitsbelow 0.1-micron line rules. Even shorter wavelengthcoherent radiation is proposed using high energy free-electron lasers. One such proposal would use theSLAC linac to generate 1-Angstrom level or ~ 10 keVphotons. Beyond this, ultrashort pulses of gammarays can be produced by backscattering intense lasersfrom relativistic electron beams. Photon energies upto 29 GeV have been demonstrated using thistechnique, opening the prospect of ultrafast nuclear orhigh energy science!

II.A.2. Frontiers and Physical Limits of SolidState Lasers

II.A.2.1 Laser-Based Mechanisms

Electromagnetic fields are proportional to the squareroot of the intensity in the focus of a coherent source.They are thus limited by the laser power (W) and thefocal size, which, in turn, depends on the wavelengththrough diffraction limits.

The laser power is theoretically limited by the saturationfluence and the pulse length, both being ultimatelydetermined by material constants (cross sections andbandwidth). In practice, additional properties likespectral absorption, thermal conductivity, temperature-and intensity-dependent index of refraction play acrucial role, limiting the number of suitable high-intensity laser materials to a handful of host materials,glasses and crystals. They typically contain a smallconcentration of ions with mostly vibrationallybroadened transitions in the near-infrared as activecomponents (solid state high power lasers).

Extensive and systematic screening of possible newmaterials over the last 5 to 10 years has yielded limitedsuccess in pushing the theoretical physical limits ofintensity and field strength. We note that at least thebandwidth of the best known materials (likeTi:sapphire) is already within a substantial fraction ofthe photon energy, which is the ultimate physical limit,and can only be substantially improved by going tomuch shorter emission wavelengths.

The physical limits of laser field strength and intensitywere increased by more than four orders of magnitudethrough introduction of chirped pulse amplification(CPA). We assume that this technique and theproperties of presently known materials willessentially determine the maximum achievableintensities and field strengths for solid state highintensity lasers over the next 5 to 10 years.

Under this assumption, the maximum power out of 1cm beam diameter of a solid state laser is between


23 Panel B

100TW and 1 PW, the intensity (at wavelengthsaround 1um and nearly diffraction limited) is of theorder of 1022 W/cm2, the electric field strength is ofthe order of 1012 V/cm (1000 a.u.). A beam diameter(limited by materials processing) of the order of onemeter would increase the focused intensity by anotherfour orders of magnitude (1026 W/cm2), and the fieldstrength by two. However, we do not believe that thisis a realistic assumption. Furthermore, even to reachthe maximum values for 1 cm beam diameter requiresenormous technological advances in laser etchnology(B- integral, stretching factors etc.). Hence, even forthe visionary purposes of the present study weanticipate the following frontiers:

Laser power: 10 petawatts (1016 watts)pulse duration: 5 fswavelength: 0.8 to 1 µmIntensity: 1023 W/cm2

Electric field strength: 1013 V/cm (2000 a.u.)Magnetic field strength: 106 Teslaponderomotive energy: ~ 10 GeV (Relativistic

scaling)

II.A.2.2 Other Mechanisms to Create Ultra-High Fields

The upper limits of the electromagnetic field strengthas well as the lower limit for pulse durations will beencountered in relativistic heavy ion collisions. At theLHC at CERN, Pb82+ ions at energies of ~3000 GeV/nucleon will collide head-on and generate electric fieldstrengths exceeding 1023 V/cm for pulse durations ofonly 10-28 s.

II.A.3. VUV- and XUV- Sources for HighIntensities and High Fields

The physical limitations of present technology invokesthe need for developments beyond our conventionalapproach for generating coherent high fields and ultra-short pulses at short wavelengths. The issue is an oldproblem associated with materials, e.g. absorption, asone pushes the wavelength beyond the VUV region.The options are either the use of nonlinear up-

conversion of some longer wavelength driving field orthe direct production of coherent light via freeelectrons either confined to a plasma or fourthgeneration accelerator based device.

II.A.3.1 Direct Frequency Conversion of LongWavelength Lasers

The following assumptions are used to estimate theultimate limits of creating ultra-high fields at shortwavelengths by direct frequency conversion of ultra-high intensity lasers. Direct frequency conversionschemes are High Harmonic Generation, the VUV-emission of relativistic free electrons under theinfluence of an ultrahigh laser field, and Comptonbackscattering from high energy electrons.

II.A.3.1.1 HHG

The study of the interaction of atoms/molecules withintense laser fields has resulted in the understanding ofthe dynamics associated with the production of highharmonic radiation in gas jets. The refinement ofexperiment and theory have led the way forunderstanding not only the single atom response to thelaser field but also the macroscopic propagationassociated with the harmonic source. Current sourcescan produce coherent radiation out to 2 nm and abrightness exceeding third generation acceleratorbased light sources. The intensity produced byharmonic generation have thus far has been limitedmore by macroscopic target considerations then thefundamental bounds placed by the single atom-laserinteraction.

If the technical constraints of the target can beovercome by some means of a quasi-phased matchedextended source of atoms,ions and clusters then it ispossible to estimate obtainable intensities. Assume anear visible drive laser of 1 PW peak power producingharmonic photons with a 1 keV limit and 10-6 powerconversion efficiency. Using multilayer mirrorsproducing a spot size of 10-3 of the fundamental beamyields an intensity of 1022 W/cm2, well in excess of anatomic unit of field. Furthermore, temporal


Panel B 24

compression of the harmonic process could produce anadditional factor of ten in the peak power, although thefundamental limit will be one optical cycle of the fieldor 20 as.

II.A.3.1.2 Free Electrons in Intense Laser Fields

Free electrons in intense laser fields undergo periodicponderomotive oscillations. At intensities above 1018

W/cm2 for visible-frequency lasers, this motionbecome relativistic. This has obvious importance forlight-source generation, and has been explored in thecontext of interactions with relativistic laser beams andwith electrons in plasmas.

The oscillating fields in lasers make them interestingcandidates for high periodicity insertion devices inaccelerators, and one could imagine such devicesproducing tailored vacuum-ultraviolet or x-rayradiation sources, possibly with high coherence.

In this context, we note several recent reseach projectsin this area. (This is NOT a comprehensive list!)

Transverse Thompson scattering of photons from anultrafast laser scattering from a linear acceleratorelectron bunch at the ALS has produced ultrafastpulses of hard x-rays.

A SLAC proposal would turn part of the linac into aFEL producing 1-10 angstrom x-rays.

SLAC has also produced 29 GeV gamma rays byCompton backscattering of an intense laser from the46 GeV electron beam. Compton backscatteringproduces extremely intense, ultra-short radiation withassociated high field strengths, which will be of maininterest to the high energy physics community. Whileit is outside the scope of AMO applications, itrepresents one of the interdisciplinary links within theDOE program.

A wakefield acceleration experiment at Michigan hasproduced MeV electrons in sub-picosecond bunches,which could produce ultrafast x-rays via bremstrahlung.

II.A.3.2 VUV- and XUV-Sources and Amplifiers

An intriguing new vista of TW, short wavelengthsources have been discussed in the context freeelectron lasers. The basic principle relies upon theefficient exchange between the photon field andrelativistic electrons accelerated in the magnetic fieldof a wiggler. In this regard, two approaches are underinvestigation each with their own merit but based on asimilar principle of high gain amplification. TheSASE (self amplified spontaneous emission) scheme isbased on the exponential amplification of radiationwhich builds up out of the noise in the first fewsections of the wiggler. Thus, the technical limit of thisemission will be based on electron beam energy andquality, e.g. emittance.

Both SLAC and the TESLA project at DESY haveproposed operation at 1 nm with an intensity of 1020

W/cm2. In the beginning of calendar year 1997,important proof-of-principle demonstrations of theSASE mechanism have been conducted at three majorlaboratories (Brookhaven, Los Alamos, UCLA) atinfrared wavelengths. These results provide animportant benchmark for testing the theoretical modelswhich are being used to scale to shorter wavelengths.Current analysis have been promising.

A second approach (BNL and Jefferson Labs) is basedon harmonic up-conversion of a laser-seeded FEL. Theadvantages of this scheme over SASE are in thecoherent properities, peak power and pulse duration(as short as 3 fs) of the output light. However, theSASE scheme is able to produce light at shorterwavelengths (below 10 nm).

Based on recent current designs in the area of VUV-free electron lasers and X-ray lasers we estimate thefollowing limits of intensity and field strength atwavelengths between 1 µm and 1nm:

@ 200nm (TJNAF FEL) design intensity 1017 W/cm2

@ 100nm (BNL FEL) design intensity 1020 W/cm2


25 Panel B

@ 10nm (TTF FEL) design intensity 1017 W/cm2

@ 1nm (TESLA FEL, SLAC FEL) designintensity 1020 W/cm2

@ 1nm hypothetical x-ray laser (10 mJ / 1 ps)intensity 1024 W/cm2

(Note that these are “back of the envelope” upperlimits which don’t take into account possible sub-optimal focusing characteristics in these sources.)

Conclusion: Current technology allows the assumptionthat the wavelength region between 1 µm and 1nm willbe covered by various high-intensity (1020 W/cm2),high-field (1011 V/cm) light sources, available to AMOphysics.

II.A.4. Infrared Sources of High Intensities andHigh Field Strengths

While today’s most intense solid state lasers operate inthe near infrared around 1 µm, no materials orconcepts are presently known to create coherent lightat longer wavelengths (10 to 100 µm) with comparableintensities and field strengths. This is in remarkablecontrast to the short wavelength region (1µm to 1 nm),where both frequency conversion methods andindependent light sources exist at intensities around1020 W/cm2 and field strengths around 1011 V/cm.

From a physical point of view one expects lowerintensities and field strengths for the followingreasons:

a. The amount of energy stored in a given volumeof an active laser medium of density rhodecreases inversely with the wavelength of theradiation.

b. The maximum bandwidth (ultimately limited bythe transition energy) also decreasesinversely with the wavelength, thus lengtheningthe pulses and decreasing the maximum power.

c. The focal spot size increases proportional to thewavelength, thus decreasing the maximumintensity. Hence, for comparable volumes of theactive medium the maximum intensity isexpected to decrease at least with the fourthpower of the wavelength.

II.B. Harnessing Quantum Dynamics

In this section, we would like to project a vision ofharnessing quantum dynamics as an importantenabling capability for the future. Here we have inmind not simply new probes of matter (be they opticalor otherwise), but more generally, the ability tomanipulate complex quantum systems to do ourbidding with a degree of dexterity that is onlyglimpsed by current capabilities. Because the systemsof interest are bound on energy scales given byelectron volts over distance scales of Angstroms, thetime scales implicit for these investigations are in thefemto-second domain and beyond, and the intensitiesinvolved are greater than 100 terawatts per squarecentimeter. Hence, a necessary component of ourvision is the need for the advancement of technicaltools to a frontier well beyond current capabilities.

As pulsewidths have decreased and peak power andwavelength control has increased in the recent decade,it has become possible to conceive of a new regime oflight-matter interaction in which the light is no longertreated as a passive observer. We are approaching anera in which we can “sculpt” dynamical systems to alarge number of purposes, improving techniques forchemistry, fast microelectronics, opticalcommunication, and even energy deposition in highdensity plasmas. Some tantalizing recent workpointing to this new area is the coherent control ofBragg scatterers at synchrotrons, which could providea way to make high-brightness sub-picosecond x-raysources (Larsson, Z. Chang, E. Judd, P.J. Schuck, R.W.Falcone, P.A. Heimann, H.A. Padmore, H.C. Kapteyn,P.H. Bucksbaum, M.M. Murnane, R.W. Lee , A.Machacek, J.S. Wark, X. Liu, B. Shan, Optics Letters22 1012-1014 (1997)); production of wakefieldplasmas for compact particle accelerators and ultrafast


Panel B 26

sources of gamma rays and relativistic electrons (R.Wagner, S.Y. Chen, A. Maksimchuck, and D.Umstadter, PRL 78, 3125 (1997)); production ofunique forms of radiation, such as “half-cycle pulses”(D. You, R.R. Jones, P.H. Bucksbaum, and D.R.Dykaar, Opt. Lett. 18, 290 (1993)); and tailoring ofcoherent xuv radiation from high harmonics (I.P.Christov, M.M. Murnane and H.C. Kapteyn, PRL 78,1251 (1997)).

Each of these has important applications to basicscience and to energy research; each also presentsspecial opportunities for contact to other areas ofscience and technology.

II.B.2. Detailed Manipulation of QuantumSystems

The tools are now at hand that one can begin to realizethe long-sought goal of reaching inside of a quantumsystem and manipulating it with the same ease that onecan manipulate a macroscopic system. Of course, theterm “quantum systems” encompasses an enormousrange of topics, and there is no one set of tools andmethods that will allow manipulation of all of them.At present there are some particular systems that arepromising for achieving this detailed level of control.

Among these are the control of the center-of-massdegree of freedom of atoms and ions in traps; thecontrol of the vibrational degree of freedom in amolecule, and the internal electronic degree offreedom in an atom.

The most important applications that will be availableif this capability is realized are no doubt yet to beconceived, but some are obvious: the control ofchemical reactions, the production of materials withunique optical and electrical properties, and thedevelopment of better understanding of quantumdynamics. There are less obvious possibilities ofsome promise including massively parallel computing,and secure communications.

II.B.3. Pulse shaping and adaptive control

II.B.3.1. What and Why

Behavior of quantum systems is determined by thetime-dependent Hamiltonian and initial conditions.Therefore, modifications of this behavior are madepossible by the application of external fields thatchange the Hamiltonian and/or the initial state of thesystem. Developments in optical technology havenow made it possible to envision detailedmanipulation of quantum systems using carefullyshaped electromagnetic pulses. This capability willallow the development of technologies at an entirelynew sub-atomic scale, and at the same time allow theutilization of uniquely quantum mechanical propertiesfor practical application. There are classical analogiesto certain types of manipulation. For example, fieldsmight push an electron in a certain direction within anatom, deposit energy in a specific chemical bond, orinduce directed electron currents in solids. However,there is also an inherent complexity in these systemsdue to their quantum nature. While this complexitycan make simple classical manipulation more difficultor impossible, it also provides additional degrees offreedom for non-classical control. For example, statesof multi-electron atoms and molecules are naturallycorrelated and entangled. These uniquely quantumfeatures are exactly what is needed for applications inquantum computing, secure encryption, and otherareas not yet envisioned that rely on the ability ofquantum systems to explore simultaneously manyparallel paths.

II.B.3.2. What do we need?

To realize arbitrary manipulation, tools are needed toshape or mold the quantum wavefunction. These toolsare electromagnetic fields and the interactionHamiltonian. Wavefunction “sculpting” has two basicrequirements. First, the amplitude of theelectromagnetic field must have temporal variationsover periods comparable to the characteristicdynamical time-scales of the system. Second, the fieldamplitude must be sufficient to cause a significantchange of the quantum mechanical state. Therefore,


27 Panel B

the development of sources that can provide intense,coherent, broad-band radiation is a prerequisite toquantum manipulation. Furthermore, optimal controlcannot be achieved without the ability to arbitrarilyalter the time- dependent structure of the fields thatare used to sculpt the quantum state. In a very realsense, the prospects for quantum manipulation are tieddirectly to our ability to alter coherentelectromagnetic fields at will. In turn, exquisite fieldmanipulation requires currently unrealized control ofthe properties of optical materials.

II.B.3.3. Problems to be solved

Incremental progress is currently being made inmanipulating dynamics in atomic, molecular, andcondensed matter systems. However, the ultimatecontrol described above cannot be realized untilsolutions to extremely challenging problems havebeen found. First, new laser and electro- opticmaterials must be developed to facilitate theproduction and control of the requisite fields. The newfrontiers for increasing the coherent bandwidth,frequency, and intensity of laser sources is discussedin Section III A above.

Second, in most systems of interest, the complexity ofthe Hamiltonian and number of degrees of freedom inthe system prohibits the a priori determination of themost appropriate time-dependent field. Therefore,active feedback techniques must be developed toadjust the field characteristics until the final quantumstate converges to the desired configuration.Implementation of adaptive feedback requires thatseveral obstacles be overcome. First, methods fordetailed control of the time-dependent fields must bedeveloped. This is no small task considering that insome situations sub-femtosecond temporal structurewill be required. Second, experimental methods fordetermining the time-dependent quantum state of thesystem must be discovered and implemented. Suchtechniques will require high spatial and temporalresolution in each dimensional coordinate and willnecessarily rely on the generation of tailored lightpulses. Third, the efficient algorithms that are

currently used in classical control problems must beadapted to specific multi-dimensional quantumproblems. These algorithms will rapidly interpret thequantum state feedback and produce an error signalfor input into the field shaping device.

II.B.3.4. (Im)Possible Dreams

In the following subsections, we outline potentialapplications of quantum manipulation to problems offundamental scientific, and possibly practical, concern.We do not imply that this is an exhaustive list.Because the realization of these dreams is inexorablylinked to future technological developments, wecannot provide specific example systems orexperimental details.

II.B.3.4.1. Controlling Coherent Collisions

At first glance, the random nature of collison processesseems at odds with the concept of a coherent collision.However, there does not seem to be any fundamentalphysical restriction which would prevent the use ofcoherent radiation to affect collision dynamics. Forexample, tailored light pulses could be used toenhance or inhibit particular photo-assisted reactionsat specific times. Alternatively, coherent fields couldbe used to produce an initial atomic or molecular statewhich is transparent to collisions at specific energiesand/or the relative orientation of colliding particles.Conversely, the atomic or molecular wavefunctioncould be manipulated so that it reacts at specifictimes, and only in certain reaction coordinates. Lastly,dynamics established in the internal degrees offreedom of an atom or molecule, A, might betransferred to a collision partner, B. In this case thecollision dynamics would facilitate excitation ofinternal direct application of radiation fields.

II.B.3.4.2. Quantum Encoding

Secure communication is an important and ongoingproblem. The properties of quantum systems makethem potential useful for encryption applications. Forexample, information (phase and amplitude) can beencoded into the wavefunction of a complicated


Panel B 28

quantum system with many entangled degrees offreedom. After a very short time, phase evolutionmakes the code unrecognizable. A time-dependentfield applied to the system at any time after thecreation of the coded state acts as a “key” whichproduces an echo of the initial state. The key can havelimited access and be universal so that it is capable ofretrieving any code stored in that system.

II.B.3.4.3. Single Atom/Molecule Electronics

Tailored light pulses produce a coherent quantum statein an atom/molecule/nanostructure. This quantumstate acts as a transistor, switch, or logic gate. Digitaldata or analog signals are input to the device using anadditional electromagnetic pulse. A third pulseinterogates the quantum state of the device providingthe readout. The complexity of device function isdetermined by the complexity of the initialconfiguration field as well as the I/O pulses. In fact,single function quantum logic gates have already beenproduced using trapped ions

II.B.3.4.4. Bond-Selective Chemistry ThroughCoherent Excitation

There has been recent progress towards this long-discussed goal of physical photochemistry. Multipleinterfering pathways toward specific ionization ordissociation products has been demonstrated (L. Zhu,et al., Science 270, 77 (1995)). Ultrafast coherentwavepacket control in molecules has also beenobserved (B. Kohler, et al., PRL 74, 3360 (1995)).Since most interesting photochemistry involvessystems in contact with rapidly dephasing media suchas solvents, a great premium is placed on ultrafast,coherent, and strong fields that can perform theirquantum-mechanical tasks in only a few tens offemtoseconds, before dephasing of the system.Furthermore, since the Hamiltonian for these systemscan be quite complicated, and only knownapproximately, adaptive control strategies will be quiteimportant.

Coherent control in the strong field limit may solvetwo fundamental problems in bond selectivechemistry: dephasing, and low yield. This is becausestrong fields can dominate the energy scale in theproblem as they reshape the molecular system. Forrecent work in this area, see G.N. Gibson, M. Li, C.,Guo, and J. Niera, Phys. Rev. Lett. 79, 2022-2025(1997), and E. Constant, H. Stapelfeldt, and P.B.Corkum, Phys. Rev. Lett. 76, 4140-4143 (1996).

II.B.4. High Intensity Optical Channeling.

When high intensity light propagates through atransparent medium, its presence alters the opticalproperties of the medium. The resulting self phase-modulation and self-focusing have already foundmany applications in short pulse laser technology,micromachining and optical propagation. Othersimilar phenomena are electromagneticaly inducedtransparency and “photon bullets,” which are stableconducting plasma channels. The latter have beensuggested for remote lightning protection.

As the power scales up, the nonlinearity changes insome important ways. Above a critical whole-beampower threshold, the principal nonlinear self-focusingmechanism becomes relativistic self-focusing. Theconsequences of this are pretty startling: An intenselaser beam focused into a plasma in this regime formsa cavitation channel, a region of low or even zeroelectron density, in which the laser can penetrate forlong distances. The channel formed in this way canbe tailored by shaping the intense optical pulse tooptimise a number of physical properties, such as thedensity, width, length, or field gradients.

One important application of this technology istransmission of the bulk of the laser energy deep intoa plasma, where it can be deposited at a plasmadensity gradient. Such schemes may be critical forefficient energy conversion using laser-compressedfusion targets (ref: Fast Ignitor workshop).

As important as this is for future energy development,it is only a small part of the potential applications of


29 Panel B

high intensity laser-channeling. This expulsion ofelectrons coupled to relativistic self-focusing thenproduces a large (GeV/m) longitudinal wake fieldwhich can accelerate bunches of electrons to multi-MeV energies over millimeter distances. Such beamshave already been shown to have some uniqueproperties, such as short bunch length and narrowdivergence.

Applications to the needs of traditional particleacceleration are obvious: the tremendous increase ofthe field gradient in these plasmas can lead to greatscaling advantages for small accelerators. There maybe some unique opportunities here as well to takeadvantage of the novel conditions in the plasmas. Forexample, the bremmstrahlung radiation producedwhen these ultrashort electron bunches hit a solidtarget may provide the best short pulse technology forultrafast gamma radiation.

II.C. Matter in Extreme Conditions

Advanced laser and particle sources providing short-pulse and high-intensity photon, electron, and ionbeams are expected to become primary enablers ofAMOP science and technology in both the near anddistant future. Each source has spatial, temporal andspectral characteristics that are unique andcomplimentary. They will allow controlled depositionof energy into various systems, producing statespreviously inaccessible in earth bound laboratories.They will permit studies of the atomic physics ofmatter under extreme and unusual conditions, whichinclude: extreme pressure (Gbar) extreme fields (E >>1 keV/Angstrom, B > 1 gigagauss)

II.C.1. Extreme Pressure

Everything from Fermi-degenerate gases,nonequilibrium plasmas and highly perturbed atomicsystems can be studied in these interactions. Extremepressures are found in astrophysical objects, pulsedpower machines, weapons, advanced x-ray sources,and ICF fusion pellets. They may also enable newroutes to IC fusion such as the fast ignitor. The

transition of the atomic structure of matter from itsdifferent states (eg. from solid to liquid to gaseousphase can be observed with unprecedented temporalresolution. Matter can also be heated isochorically(without change in density) to extreme temperatures(kilovolts). These studies are important for theunderstanding of radiation damage, material ablationfor thin-film deposition, fusion, pulsed power,propulsion, waste disposal and plasma processing.The x-ray emission from the laser-heated solid-density plasma also makes a bright, ultrafast andcoherent table-top x-ray source that can be used tostudy transient phenomena on an ultrashort timescale.Such a source can also be used to study nonlinearoptics with inner-shell transtions. It is relevant to x-ray lithography and indirect drive ICF. In the lattercase, the source can be used as an imager to freezehydrodynamic motion and provide a test bed for thestudy of material properties at a wavelength relevant tohohlraums.

II.C.2. Extreme Fields

When the work done on an electron by the field overthe distance of the laser wavelength equals theelectron rest mass (1018 W/cm2 ), we enter a newregime of nonlinear optics. In this case, a plasmamedium can be modified by the laser field in such away that it can be used to guide the laser light at highintensity over a much greater distance than it would invacuum. The light pipe that results when theelectrons and ions are driven from the axis of thechannel can be used to guide another laser pulse. Thisis the basis for the proposed fast-ignitor fusionconcept. Gigagauss magnetic fields and electron-positron plasmas may be created. Interesting ionicstates may be created and relativistic effects dominatethe interaction physics. The largest electrostatic fieldsever produced in the laboratory are generated inwakefield plasma waves, which can accelerate usefulnumbers of electrons to gigavolt energies in a distanceof just a centimeter. This is a new regime ofparameter space that may result in new properties ofmatter and is relevent to advanced accelerators,electron sources and fusion energy concepts. One of


Panel B 30

the exciting possibilities opened by high-intensitylaser and ultra short relativistic electron and ionbeams is the creation in the laboratory of extremelyhigh electric and magnetic fields at the criticalCompton limit of 1016 V/cm. In these extreme fields,the relation between energy and the vacuum aremodified in a dramatic way. One expectedconsequence is the breakdown of the vacuum. Inaddition, the large energies deposited over amacroscopic volume would be sufficient to generatecollective and coherent effects. This will result in theformation of an e+e- plasma. Just as one needs atheory of non-linear optics to understand strongelectromagnetic interactions, a theory of non-linearQED will be needed to understand these collectiveeffects.One concept for achieving these fieldscoherently backscatters an intense femtosecond laserfrom relativistic electron beams to generate fields onthe order of 1029 W/cm2. Coherent effects in thebackscattering would then enhance the field byne2.Also, laser-based sources can be built for thegeneration of energetic electrons (laser accelerators),ions or neutrons. The pulse duration of these electronbunches can potentially be subfemtosecond, which issufficiently short to lead to novel effects. For instance,coherence effects can enhance the intensities when theelectrons in the beam are acting collectively (scalingas N2 instead of N). The duration can also becomeshorter than several interesting characteristictimescales, such as Auger lifetimes.

II.C.3. Connections to Other Fields at DOE

High field AMO physics experiments are related to thenuclear physics studies of interactions betweenrelativistic heavy ions at RHIC. In contrast to theinteractions at RHIC, which involve quarks andgluons, the AMO experiments described here involvephotons and leptons. Studies of the interaction andcreation of particles in these experiments may be asubset of those observed at RHIC. Because there arefewer types of interactions interpreting theseexperiments maybe be less complex. In addition, theintense fields found in pulsars may be recreated byhigh power laser lasers (or other sources). In these

intense fields, atoms may undergo recombinationwith electrons or other particles that are pulled out ofthe vacuum.

Panel C: Surface Interactions With Photons,Electrons, Ions, Atoms, and Molecules

Chair:J. Wayne Rabalais (University of Houston)

Panelists:Raul A. Baragiola (University of Virginia)Barbara H. Cooper (Cornell University)

F. Barry Dunning (Rice University) Charles S. Fadley (Lawrence Berkeley National Laboratory)

J. William Gadzuk (National Institute of Standards and Technology)Wilson Ho (Cornell University)

Dennis C. Jacobs (University of Notre Dame)Theodore E. Madey (Rutgers University)Steven J. Sibener (University of Chicago)

Nicholas Winograd (Penn State University)


Panel C 32

Executive Summary

NEEDS: The present understanding of the interactionsand behavior of particles (ions, atoms, molecules,clusters, electrons) and photons with surfaces at thebasic atomistic level is far from complete. A largenumber of outstanding questions remain that pertainto, for example, the detailed dynamics of particle-surface interactions, localized energy deposition atsurfaces, nonequilibrium phenomena, and theproperties of atoms and molecules on a surface.Central to improving the present understanding iscontrol of the energetic, spatial, and temporal degreesof freedom of the particles involved in the interactions,coupled with manipulation of atoms and molecules atthe surfaces. Improved understanding of surfacephenomena is fundamental to the development andsynthesis of materials with specific physical andchemical properties. This underpins futuretechnological advances in areas such as energyproduction and conservation, microelectronics, and theenvironmental and life sciences. Many of theseapplications are based on phenomena that occur atsurfaces and interfaces, e.g. solid/vacuum, solid/gas,liquid/vapor, and solid/liquid, as a result of stimulationby particles or photons. Studying such phenomena hashistorically been difficult and challenging. Althoughmuch progress has been made to date, the microscopicunderstanding of many interfacial phenomena remainsbelow that currently existing for homogeneous phases,i.e. gas-phase atomic and molecular physics and solidstate physics.

OPPORTUNITIES: A unique opportunity currentlyexists to advance this challenging area of basicscience. This involves the integration of the expertiseand technology from atomic, molecular, and opticalphysics (AMOP) with that of surface science. AMOPscientists have developed an in-depth understanding ofthe properties and behavior of gas phase atoms andmolecules, highly sophisticated experimentaltechniques to produce and analyze particle beams inwell controlled states, and advanced theoreticalmethods. Surface scientists have developed a

comparable battery of techniques for control of thestate of a surface and improved theories, particularlyfor thermal processes at the gas/surface interface. Therelatively recent advances in the production of largeclusters of atoms and molecules represents an area thatis naturally intermediate between AMOP and surfacescience. Each area has much to learn from theexpertise of the other. For example, for almost everyparticle-surface interaction, there exists a comparableanalog in gasphase collisions.

BENEFITS: This synergistic relationship betweenAMO and surface scientists will enable new advancesin:

• experiments that involve selection, control, non-thermal chemical manipulation, and completeanalysis at the atomic and molecular level.

• theory that couples the discrete excitations ofatoms and molecules with the continuumexcitations in solids.

• the use of molecules and clusters for achievingan atomic level understanding of themechanisms responsible for chemical catalysisand gas-surface energy transfer.

• novel schemes for materials growth andnanostructure fabrications.

The synergy will facilitate progress in the importantareas of: (1) dynamics of excited states at surfaces, (2)non-equilibrium phenomena induced by particles andphotons at surfaces, (3) control and manipulation ofparticle- and photon-surface interactions at the atomicand molecular level, and (4) advanced techniques forprobing particle- and photon-surface interactions.


33 Panel C

Outline

I. Fundamental Understanding of Particle- andPhoton-Interactions with Surfaces

II. Non-Equilibrium PhenomenaA. Low Density Excitations

A1. Non-Equilibrium Conditions in FilmGrowth and Modification

A2. Theoretical ApproachA3. Technological Importance

B. High Density ExcitationsB1. Dynamics of Electron TransferB2. Dynamics of Highly Concentrated

ExcitationsB3. Mechanism of Energy Flow in

CollisionsB4. Particle-Surface Interactions as Probes

of Surface Dynamic PropertiesB5. Laser Induced Non-Linear

Phenomena

III. Control and Manipulation at the Atomic andMolecular LevelA. Energetic Control and State SelectionB. Spatial ControlC. Temporal Control

IV. Other Advanced Techniques for ProbingParticle- and Photon-Surface InteractionsA. Beams of Ions, Atoms, Molecules,

Clusters, Electrons, and PhotonsB. Advanced Combinations of Supersonic

Beams, Atomic Traps, and Atom OpticsC. Femtosecond (fs) Surface SpectroscopyD. Spatial- and Temporal-Imaging of Surface

Atomic StructuresE. Coincidence TechniquesF. Surface Studies in High Pressure

EnvironmentsG. Higher-Efficiency and Higher-Speed

Detectors

V. Key Research Areas in this Report that areUnder-Represented in the United States

VI. Summary List of Basic Unsolved ProblemsInvolving the Interactions of Low EnergyParticles (Electrons, Ions, Atoms, Molecules,Clusters) and Electrons with Surfaces

VII. Summary List of Applications of Particle andPhoton Interactions at Surfaces

VIII. Conclusion Statement

I. Fundamental Understanding ofParticle and Photon Interactions withSurfaces

Considerable progress has been made over the pastseveral years in describing the properties of clean andadsorbate-covered surfaces. Less is understood aboutthe dynamical processes that occur when particles(electrons, ions, atoms, molecules, and clusters) andphotons impinge on a surface, especially when excitedstates are involved. A detailed knowledge of theformation and relaxation of excited states is ofprincipal importance in understanding nonequilibriumprocesses and in exploiting these processes in novelapplications.

Excited electronic states are formed near the surface asa result of the absorption and scattering of photons andelectrons and by the impacts of ions, atoms, molecules,and clusters. Energy is also deposited directly intonuclear motion by momentum transfer to surfaceatoms. These impact-induced excitations range from,for example a change in the charge state of a low-energy scattered projectile or an electron-hole paircreation by the absorption of a photon, to high densityexcitations such as track formation and materialablation resulting from heavy or highly charged ionimpact. State-of-the-art experiments utilizing surfacesthat are well-characterized, i.e. atomically clean,aligned structures, and highly controlled beams ofparticles that are mass-, energy-, charge-, and state-selected, collimated, and pulsed (in some cases), inconjunction with techniques for complete delineationof the interactions in terms of elemental-, spatial-,


Panel C 34

temporal-, and final state-resolution, are needed inorder to reach a new level of fundamentalunderstanding.

Beam-surface experiments that probe the fate of ascattered projectile or that follow the nascentexcitations in the surface are required to provide acomprehensive understanding of the relevant energyand charge exchange mechanisms and the relaxation ofexcited states. For example, final electronic statedistributions of scattered atoms or molecules, coupledwith theoretical models of nonadiabatic electronicinteractions, reveal the charge transfer mechanismsand information about the lifetimes of excited statesnear surfaces. Desorption of atoms or moleculesfollowing electron impact indicates the coupling ofnuclear motion to electronic excitations. Ultra-lowand thermal collisions of neutral atoms and moleculeswith interfaces are also of high interest, with suchgentle interactions dominating phonon exchange,sticking, and chemical reactivity. Fragmentationpatterns following cluster impacts reflect energyredistribution pathways and the time scales over whichenergy redistribution occurs. Electron emissionsubsequent to impact identifies the specific excitedstates produced in the projectile and/or substrate.Moreover, time-resolved spectroscopies involvingultrafast lasers or third generation photon sources willallow one to directly follow how electronic excitationscouple to other degrees of freedom and how localizedexcitations decay in space and time. Experimentaladvances, focused on increasing control of the incidentspecies and improving temporal and spatial resolutionfor probing the final states, will provide new windowsfor understanding dynamical pathways.

New theoretical developments are essential toaccurately treat the inherent many-body nature ofparticle-surface interactions, including electronicallyexcited states at surfaces and relaxation pathways.One of the challenges is the treatment of phenomenathat evolve over a time scale of some ten orders ofmagnitude. A simplistic delineation is as follows: (1)The collision regime (< 100 fs) involving direct

nuclear collisions, excitation of electronic transitions,creation of a local non-equilibrium, high-energy-content region (mini-collision cascade), and generationof vacancy and interstitial pairs. (2) Thethermalization regime (0.1 - 10 ps) involvingdissipation of excess energy through phonons, rapidquenching of the cascade region, and materialrecrystalization. (3) The diffusion and recombinationregime (1 ns - 10 µs) involving thermal diffusion andannihilation of point defects by diffusion torecombination sinks. Meeting this theoreticalchallenge requires well coordinated expertise frommany areas in AMOP as well as from chemical andcondensed matter physics. Theoretical modeling andsimulation of the time evolution of an excited atomicor molecular system coupled to a surface (with all itsunique attributes), possibly in an intense AC field(laser) or DC field (scanning tunneling microscope(STM), tunnel junction, field ion or emission tip),requires creative new approaches transcending bruteforce diagonalization of large matrices or numericalsolution of large but finite systems of ordinarydifferential equations which incorporate standard forcelaws or manageable potential energy surfaces.

Because of the importance of electron chargeredistribution and transfer, quasi-localised resonancestate formation, and field-induced electron dynamics,the coupled, multi-dimensional quantum equations ofmotion must be able to treat the electronic and nucleardynamics on an equal footing. An additional challengearises because the solid and laser fields each present acontinuum of excited states that must be consideredand exploited within theoretical models. Low andultralow energy interactions present another excitingchallenge due to the unusually large deBrogliewavelengths and sharp resonances which characterizesuch interactions. Surface charge density distributionsand collective surface phonon modes must receiveaccurate treatment if such calculations are torealistically model atomic scale behavior. It isenvisioned that innovative combinations ofappropriately statistically averaged classical (Newton,Langevin equations, etc.), semiclassical (WKB,


35 Panel C

Gaussian wavepacket dynamics, etc.), and quantum(coupled channels, wavepacket propagation on grids,soluble model Hamiltonians, etc.) particle-field modelswill be required, together with intuition fordifferentiating the relevant from the irrelevant aspectsof the interacting system. This work should provide atractable, useful, and informative theory onfundamental processes at surfaces that can serve toguide, interpret, and predict experimental results.Close experimental-theoretical coupling must bemaintained, each helping to guide the other in this“self-correcting” drive to achieve a fundamentalunderstanding of the interactions involved. This, inturn, will lead to unprecedented control andexploitation of these interactions in importantapplications such as nanoscale surface modification,low temperature growth of novel materials, andcreation of metastable phases.

II. Non-Equilibrium PhenomenaEnergetic particle and photon beams provide a meansof producing conditions in the surface region that arefar from thermodynamic equilibrium. Understandingthe fundamental particle- and photon-surfaceinteractions will improve our ability to exploit them.The new degrees of freedom resulting from the use ofcontrolled beams allow non-equilibrium excitation ofsolids, which can be used to deposit films, synthesizenew metastable materials and phases and understandradiation effects. This section is roughly divided intotwo subsections, i.e. low density and high densityexcitations.

A. Low Density Excitations

By low density excitations we refer to those created inthe near surface region by, for example, the impacts ofrelatively low energy atoms, molecules, and ions.

A1. Non-Equilibrium Conditions in FilmGrowth and Modification

Particles of a wide range of incident kinetic energiescan selectively influence film growth chemistrydepending on the nature of the potential energy

surfaces which govern such processes. Consider, forillustrative purposes, the energy regime that is directlyapplicable to film growth and modification on surfacesinvolving appreciable entrance channel barriers. Inthis instance, a reasonable lower limit is a kineticenergy of the order of bond energies, ca 1 eV. Theenergy released upon bond formation is dissipated asphonon-like excitations and lattice distortions whichperturb the local environment in a manner similar tothe excess kinetic energy supplied by the incidentbeam. At this lower limit, chemical bondinginteractions become significant, the binary-collisionapproximation becomes questionable, and inelasticinteractions between the incident particle and substratecan alter the incident trajectories. A reasonable higherlimit, in the case of ions, is a kinetic energy of theorder of a few keV. In this range the sputtering yieldbecomes equivalent to, or higher than, the beam flux,classical ion-trajectory simulations using a binary-collision approximation provide a satisfactorydescription of many of the collision phenomena, andthe impinging ions remain in the near-surface layers.The thresholds for penetration, displacement,sputtering, and reaction occur in this energy range. Inthe case of neutral atoms, supersonic beams can easilyspan the range from sub-thermal energies up to severaleV. This allows one to selectively and systematicallyexamine chemical reaction channels which becomeaccessible at ever increasing energies. Some of thesechannels may lead to new or more highly optimizedgrowth conditions than are typically accessible inCVD environments. Moreover, supersonic beamexperiments permit intentionally selected growthprecursors to impact the target sample, greatlysimplifying the analysis of growth mechanisms.The ability to select the energy, type, and arrival rateof particles, incident polar and azimuthal angles,substrate temperature, background gases, andstoichiometry allows control of the penetration anddefect energy thresholds, film growth mechanisms,isotopic composition, low temperature epitaxialgrowth, chemical reaction mechanisms, collision-induced surface chemistry, energy flow in collisions ofmolecular species, and coupling of gas and solid phase


Panel C 36

excitations. While conventional film growthtechniques rely on high temperature to activatechemical reactions and diffusion processes, growth ofnovel high quality materials at low temperatures canbe achieved by coupling the internal and translationalenergy of the incident particles with the surface energymodes. Low temperature growth is becomingincreasingly important in a number of technologicalapplications where it is desirable to preventinterdiffusion at interfaces and to stabilize nanoscalestructures. The fundamental mechanismsdistinguishing energetic beam deposition from thermaltechniques result from localized deposition of energyby the incident particles. The energy stored in theincident particles can be used to overcome reactionbarriers and induce nonthermal chemistry.Electronically excited states of impinging atoms is anexample that is open for exploration. Nonequilibriumsurface atom configurations can change nucleation andgrowth modes, which include processes such asadatom and vacancy production and enhancedinterlayer mobility. The challenge is to learn tobalance the beneficial effects of ion irradiation, such aslocal relaxation, creation of mobile vacancy/interstitialpairs, and enhanced diffusion, with the undesirableeffects, such as permanent defect formation, latticedamage, sputtering, atomic mixing, etc. It is necessaryto understand the role of particle energy, mass, andcharge in penetration of surfaces, displacement ofatoms, stimulation of chemical reactions, productionof metastable structures and high density materials,and enhancement of epitaxy.Other non-thermal processes in surface modificationby energetic beams include electron- and photon-stimulated desorption and reactions. These processesinvolve electronic excitations that lead to bond-breaking. The manifold of possible electronicexcitations allowed in gaseous molecules may be verydifferent in the condensed phase. Many challengesremain in understanding energy transfer, chargetransfer, and the dynamics of excited-state evolution.An improved understanding of energy flow in thermalenergy collisions is also needed if we are to predictenergy transfer between impinging neutral particles

and clean/adsorbate covered surfaces. Suchinteractions, typically mediated by phonons, arecrucial to an improved understanding of surfacephonon spectroscopy, i.e. bonding at surfaces, sticking,and energy accommodation in general.

A2. Theoretical Approach

A full understanding of the interactions involved willrequire the more sophisticated atomistic simulations thatare currently emerging. A central problem that needs tobe resolved is the importance of the coupling betweenelectrons in the solid and the moving atoms in thecollision cascade initiated by the energetic interactions.Molecular dynamics simulations of radiation effectshave incorporated this special case of electron-phononcoupling in a parametric form, and show thatsolidification and phase transformations in the energeticatomic cascades are affected by the uncertainty in thiscoupling. Recent attempts have focused on ab-initiodescriptions of energy levels and energy transfer inenergetic cascades, but this type of theory is still in itsinfancy. Improvements are needed in calculating thefate of reactive molecular precursors with reactiveinterfaces. Sophisticated electronic structurecalculations analogous to those developed for quantumchemistry and precision atomic physics applicationswould have high impact in this area of endeavor.

A3. Technological Importance

Many techniques employ energetic particles in thegrowth process. These include sputter and plasmadeposition, pulsed laser deposition, and ion-assistedand direct ion beam deposition. These techniquesoften involve a wide range of incident species andenergies. Experiments with highly controlled incidentbeams (ions, neutral molecules, clusters, photons) andreal-time in situ diagnostics are critically needed togain a better understanding of the role of energy inmodifying growth modes and materials properties.This new understanding will enable us to controlimportant film properties such as structure,stoichiometry, internal strain, surface and interfaceroughness, and adhesion. Energetic beams have been


37 Panel C

shown to have beneficial effects in the growth ofmagnetic multilayer materials and wide band gapsemiconductors (SiC, Group III nitrides, anddiamond). This approach shows great promise forcontrolled growth of novel materials and metastablephases. In order to achieve spatially resolveddeposition and surface modification, it is desirable todevelop new techniques relying on atomic optics andion beam focusing. Of course, energetic electron andphoton beams will continue to be of criticalimportance for lithography in microelectronicsapplications as dimensions are reduced to the sub 100nm range.

B. High Density ExcitationsAn exciting new development in the field ofinteractions of particles and photons with surfaces is toproduce extended regions in solids that are excited farfrom equilibrium conditions through the impacts ofenergetic heavy ions and slow highly charged ions,clusters or large molecules. The highly unusualconditions which result cannot be described withexisting models of particle-solid interactions.Questions of interest include the following:

B1. Dynamics of Electron Transfer

Interactions between ions and surfaces present newchallenges which arise from the many-body target andits low geometric symmetry. Well-definedexperiments involving ion neutralization and chargeexchange at surfaces are needed to probe thresholdsfor excitation and ionization, to examine energy levelshifts and broadening, and their dependence onparticle velocity, surface electronic structure, andcrystal structure and to investigate the effects ofexternal fields and hybridization of electronic statesnear surfaces. The use of highly charged ions (HCI)provides a test bed for models that go beyond theindependent particle description. HCI can range fromHe2+ to U92+ . At the upper end of this charge scale,they are a source of a wide variety of new physics.The electric field produced by a HCI and its image caninduce enormous distortions of the surface potential of

a solid, stimulating multiple electron transfers thatresult in the transient formation of “hollow atoms”.Electrons evolve in unusually high n orbits whichwould resemble Rydberg states except that they arepolarized by the huge fields and they merge intoextended electron states of the solid.

B2. Dynamics of Highly ConcentratedExcitations

Ionization spikes (dense plasma) in non-metals, theirevolution in some cases to a dense exciton gas, andtheir role in particle ejection, and non-thermaltransformations in the solid (cratering, bond-breakingand reforming, phase transitions, etc.) cannot beexplained by existing theories. These phenomena, thatcan be started by fast ionizing particles or slow highlycharged ions, are important radiation effects in livingtissue and semiconductor devices. A frontier in thisarea is the development of models for the complex,multiple-particle excited region and the time-varying,inhomogeneous electric fields. This can beaccomplished by a new generation of moleculardynamics simulations which can include theappropriate local electric fields and many bodyinteractions between ions, excited atoms andmolecules.

B3. Mechanism of Energy Flow in Collisions

The flow of energy in collisions of clusters and largebiomolecules with surfaces is a complicated, albeitvery important problem. Current atomistic andcontinuum mechanics approaches are found,particularly for large, fast projectiles, to be insufficientto explain the coupling and evolution of vibrationaland electronic excitations in the cluster and at thesurface. These interactions result in extremephenomena that include the formation of very hotplasmas in the solid, accompanied by photon andparticle emission. This can lead to shock wavespropagating through the material, cratering, andejection of material. The projectile cluster candisintegrate in the collision or, at low velocities, bereflected in an unusual excited state, which includes


Panel C 38

shape distortions, internal shocks, and vibrations. Therelaxation of the cluster can result in fragmentation andthermionic electron emission and evaporation. Thelatter permits study of highly localized thermalphenomena. A goal here is to obtain completecharacterization in terms of the underlyingmechanisms, including molecular and surface quantumstate-specific excitations, reactive processes, molecularfragmentation, and dissociative deposition.Controlling the many reaction channels is also ofinterest, as are interactions with complex surfaces(frozen gases, organics, polymers, clusters).

B4. Particle-Surface Interactions as Probes ofSurface Dynamic Properties

Plasmons, collective excitations of valence electrons,provide an example of electron correlation effects thatcan occur at surfaces or in the bulk of solids and smallparticles. A long standing fundamental problem insolid-state physics has been how plasmons form andevolve in time and space, including their decay intoquasiparticle excitations. It may soon be possible tostudy this important problem as a result of thedevelopment of new ultrafast laser techniques thatprovide a time resolution smaller than, or of the orderof, the response time of the valence electrons (10-16 -10-14 s). Surface plasmon excitations by photons andhigh energy particles can be monitored optically forreal-time characterization of surfaces during materialsprocessing and synthesis.

B5. Laser Induced Non-Linear Phenomena

Non-linear phenomena in the coupling of high powerlasers with solids can lead to ejection of material fromthe surface. These phenomena induce a wide range ofphysical and chemical processes, depending on theexcitation density, that have analogies and differencesto those caused by ions. The unusual states of matterthat are produced can be studied using AMOPtechniques, such as picosecond (ps) lasers, sub-nsparticle detectors, and coincidence techniques. Highpower lasers can be used for materials synthesis,micromachining, desorption, and other surfacemodification processes.

III. Control and Manipulation at theAtomic and Molecular Level

The previous two sections have focused onexperimental and theoretical approaches to addressunanswered questions regarding fundamental particle-and photon-surface interactions, and the non-equilibrium excitations that they entail. We havestated that an increased understanding of theseinteractions will enhance our ability to exploit them.Here we elaborate on techniques and approaches thatwill ultimately lead to our ability to understand,control, and manipulate these interactions at theatomic and molecular level.

Understanding of the physical and chemical propertiesof materials and of the means for achieving energetic-,spatial-, and temporal-control and manipulation ofatoms and molecules are the fundamentalunderpinnings upon which nanotechnology is based.Atoms and molecules impinging on a surface can nowbe prepared in specific translational, electronic (exitedions and neutrals), spin, and internal states (vibrationaland rotational), providing well-defined reagents. Onceadsorbed, selective and nonthermal chemical processescan occur, driving a diverse body of research. Thescanning tunneling microscope (STM) has become apowerful tool with the ability to probe and manipulatesingle atoms and molecules. Using femtosecond (fs)lasers, it is possible to access electronically excitedintermediates and tailor intramolecular responsedynamics. Both the tunneling electrons from the STMand photons from fs lasers can induce forcesassociated with select electronic excited states and canbe used to “push” constituent atoms along desiredreaction pathways, including motion on the surface,the breaking and forming of individual bonds, and thedesorption of products. From measurement of thefinal translational and internal state distributions of thedesorbed molecules, further insights into themechanisms driving atomic and molecular motion canbe obtained. Since charge transfer underlies many ofthe mechanisms involved in atomic and molecularmanipulation and reaction, it is important to studyelectron interactions with atoms, molecules, and


39 Panel C

clusters at surfaces. The energy levels and spatialdistribution of excited states of atoms and moleculesare directly involved in these interactions. Little isknown about these states and theoretical investigationsare critically needed for guiding the interpretation ofexperimental results and in making predictions of howindividual bonds are made and broken. For electronand photon interactions with surfaces, the goal is tounderstand the mechanisms of excitation, ionization,electron capture, desorption, sputtering, reactions,damage, ablation, and charging of insulators.

A . Energetic Control and State Selection

Materials modifications can be controlled byimpacting the surface with atoms and moleculesprepared in specific energetic states. Hightranslational energies can be used to overcomereaction barriers. Molecules in one vibrational statemay be more effective than another in breaking orforming a chemical bond. Electronically excitedatoms and molecules can also promote chemicalreactions, as is evident in plasma assisted depositionand etching. By preparing atoms and molecules inwell defined excited states, it is possible to unravel themechanisms and dynamics of many complicatedreactions in catalysis, atmospheric chemistry, andplasma reactors. State selection of atoms andmolecules offers another dimension in the control ofmaterials and the opportunity to tap into the extensivetools which have been developed by the AMOPcommunity. In addition to obtaining a basicunderstanding of fundamental interactions betweenatoms and molecules with surfaces, state selection isexpected to result in reactions which cannot beaccessed by atoms and molecules with a thermaldistribution. For example, new adsorption sites can beoccupied and reactions can occur on entirely differentparts of the potential energy surface. Species on thesurface resulting from such reaction pathways can beprobed by the STM. By creating an interferencepattern on the surface with photon and atom beams,nanoscale structures can be formed on the surfacewhich are composed of materials with novel structuresand compositions.

B. Spatial Control

Recent advances in scanning probe technology make itpossible to study single atoms and molecules,providing an unprecedented opportunity to understandand control materials. It is now possible not only toimage with atomic resolution, but to obtainspectroscopic information and to induce chemicalreactions with single bond precision. By examiningindividual atoms and molecules, it is possible to isolateenvironmental effects and obtain their intrinsicproperties. Results from such studies are particularlyattractive to theoretical investigations. Since tunnelingelectrons are confined to atomic dimensions,manipulation of single atoms and molecules ispossible, providing the opportunity of buildingmaterials from basic atomic units. Atoms andmolecules can be arranged into specific configurations,leading to nano-materials with novel properties. Theability to control and manipulate individual atoms andmolecules and to induce chemical reactions withatomic resolution is the basis for single electrondevices with ultra-small dimensions which can operateat room temperature.

Another exciting and AMOP-based opportunityinvolves stimulation of new research activities in gas-surface atom optics, incident beam control, and patternformation. The use of either physical,electromagnetic, or optical fields can lead to well-focused beam spots of selected reagents on surfaces.Standing wave-interference patterns, already in use forsuch purposes, are but one application that will comefrom this new field of endeavor. This may well lead tothe controlled fabrication of nanoscale structureshaving equivalent impact to the more traditionaltechnologies of lithography, STM-based manipulation,and self-assembly.

C. Temporal Control

Surface chemistry can be induced by tunnelingelectrons from STM, photogenerated electrons from afs laser, and particles (ions, atoms, molecules,clusters). Several different mechanisms are known tobe important in these interactions. For example, a


Panel C 40

mechanism which involves inelastic negative ionresonance scattering is often responsible for theinduced reactions. Using fs lasers, the goal is to probethe time scales for bond breaking and formation on thesurface. Since the photons are confined to a pulse of10 - 100 fs duration, the induced chemistry can besignificantly different than that induced by cw and nslaser and particle pulses. Excitation with fs laserpulses also allows the study of energy transfer andcoupling of the electronic degree of freedom to thenuclear motion. It is very useful to consider fs inducedchemistry of the same molecules in the gas phase andin solutions, which provides a general understandingof chemical dynamics. In excitation with fs laserpulses, important information comes from using anexcitation time scale that is shorter than, orcomparable to, the time scales involved in energyrelaxation and transfer. By stretching the laser pulse toa ps timescale, the excitation rate becomes smallerthan the relaxation rate, resulting in different reactionpathways. The combination of fs lasers with STMsprovides an opportunity to probe, manipulate, andcontrol matter on the fs temporal and angstrom spatialscales. Nonlinear optical excitation is expected at atunnel junction, resulting in the measurement ofatomically resolved optical properties.

IV.Other Advanced Techniques forProbing Particle- and Photon-SurfaceInteractions

In addition to those mentioned in earlier sections ofthis report, several unique new techniques have beendeveloped over the past decade, many of which arestill evolving, that offer unique opportunities forprobing particle-and photon-surface interactions.Many of these new techniques have strong overlapwith existing AMOP interests and expertise.

A. Beams of Ions, Atoms, Molecules, Clusters,Electrons, and Photons

The use of highly focused beams to study surfaces isof interest for several reasons. A focused beam can

deposit more energy or material in a small spot, thusstimulating changes in the local effective temperatureor concentration of reacting/exciting species. Suchbeams can also be rastered over the surface (either bymoving the beam or by moving the sample) so as toyield laterally-resolved images of the surface. Suchimaging is becoming increasingly more important forcomplex, technologically or environmentally relevantsurfaces that are often highly heterogeneous in thelateral dimensions. Rastering can also be used to writenanometer-scale patterns on surfaces in a variety ofmaterials, thus presenting exciting possibilities for thecharacterization and production of next-generationsemiconductor circuits and magnetic storage devices.By combining imaging while sputtering, it is possibleto carry out 3D tomography of solids of arbitrarycomposition at an unprecedented resolution.Commercially available ion beam sources permitfocusing of Ga+ ions at 15-50 keV down to a spot sizeapproaching 5 nm. With further development usingAMOP techniques, it should be possible to improve onthis resolution and/or create beams of state and energyselected ions at lower energies. The sizes of suchbeams are thus smaller than the collision cascadeinduced in the solid by the projectile itself, opening upthe possibility of studying the physics of the cascadeprocess in more detail than has been possiblepreviously. With associated theoretical interpretation,such studies would have a very positive impact on thetechnique of secondary ion mass spectrometry that isan ubiquitous tool in materials science and,particularly, in the electronics industry. These focusedprobes, when combined with mass spectrometry, offernew possibilities for high-spatial-resolution chemicalimaging with applications in geology, environmentalscience, and biology.

Photon beams from third-generation light sources canbe focused down to 10-20 nm, permittingsimultaneous spectroscopy and microscopy (viasample scanning). Such “spectromicroscopy” can befruitfully applied to atomic and molecular speciesdeposited on surfaces, e.g. from ion or atom beams.Using both focused ion beams and focused photon


41 Panel C

beams together would permit studies of the rates ofdiffusion of species as well as changes in chemicalstate with lateral displacement away from the positionof the primary collision. Core-and valence-levelphotoelectron spectroscopy, Auger electronspectroscopy, and soft x-ray absorption spectroscopycan be carried out with unprecedented resolution andexperimental control, including the variation of lightpolarization and measurement of outgoing electronspin.

Optical pumping or magnetic state selection viahexapole magnets can produce beams of aligned atomsand ions. Studies of spin dependences and theirsurface interactions will be invaluable for probing ofmagnetic surfaces and interfaces, giantmagnetoresistive structures, colossal magnetoresistivecompounds, and high-temperature superconductors.

B. Advanced Combinations of SupersonicBeams, Atomic Traps, and Atom Optics

Supersonic beams have revolutionized ourunderstanding of precision spectroscopy and reactiondynamics. Atomic traps have had a similar impact inmany fields, spanning the range from atomic fountainsfor clocks to Bose-Einstein condensation. Atom opticsare now being applied for precision measurements,interferometry, and pattern generation includingholography. It is clear that outstanding scientific andtechnological advances will result from the combineduse of these incisive techniques. New techniquesinvolving high-brightness neutral beams, spatially-guided and tightly-focused neutral beams, andmesoscopic pattern deposition will follow from suchstudies. Already extant exploratory studies involvingultra-cold and spatially controlled neutral beamsinjected into optical fibers may lead to new classes ofatomic gyroscopes. Atoms exiting such fiber-opticsbased devices may find application as atomic“fountain pens” for lithography and mesoscalechemistry.

C. Femtosecond (fs) Surface Spectroscopy

The development of ultrashort laser pulses that arenow typically 60 fs in duration, but with values asshort as 5-10 fs having been obtained, has opened upnew possibilities for studying electron, photon, andenergy transfer dynamics at surfaces. In theseexperiments, single- and multi-photon absorption canlead to time-resolved electron excitation, e.g. intoimage states near surfaces that are one-dimensionalanalogues of Rydberg states in atoms. Electronemission from these excited states can be studied byusing pump-probe techniques. Thus, the energies andlifetimes of electronic states near surfaces can beprobed as well as the influence of photochemically-induced surface reactions. Alternatively, the energyfrom a single laser pulse can be deposited over a shorttime into secondary electron creation, with theseelectrons then driving surface chemical bond changesthat can be followed in time, e.g., via time-of-flightmass spectrometry.

D. Spatial- and Temporal-Imaging of SurfaceAtomic Structures

New probes of surfaces based on ion and photonbeams also offer the possibility of studying the internalatomic structures of surfaces on time scales relevant tosurface chemical processes. For example, in therecently developed technique of scattering andrecoiling imaging spectroscopy (SARIS), the atomicstructure is determined via two-dimensional ionscattering patterns from a crystal in a time-of-flightsystem. The pattern acquisition time is in the few-second range and individual images of the scatteredand recoiled particles can be resolved on a 10 ns scale.For many surface reactions at near-UHV conditions,this is rapid with respect to surface atomic diffusionand reaction. These ion scattering images exhibitextreme sensitivity to surface structure, providing areal-space, element-specific surface crystallography.This technique also offers new opportunities in thedetection of surface hydrogen, the study of non-planarscattering events, the analysis of disorder at surfaces,determination of surface diffusion and kinetics, and the


Panel C 42

study of ion-surface charge transfer characteristics.New applications of this ion scattering technique toliquid surfaces, which are currently emerging, offerthe possibility of monitoring dynamics on fluidsurfaces. With high-brightness photon beams fromthird-generation synchrotron radiation, it is currentlypossible to measure high-resolution chemical-state-resolved core-level photoelectron spectra on timescales in the second range and to measure two-dimensional photoelectron scattering and diffractionpatterns from the crystal lattice in minutes. Similarinformation can be obtained on a smaller scale byAuger mapping with finely focused electron beams.Such ion and photon probes can follow thedevelopment of different chemical states in time, thusdirectly deriving kinetic information. The scatteringimages and diffraction patterns also permitdetermination of local atomic structures for thevarious chemical states present. Such measurementscould be applied, for example, to surface reactionscarried out with state-selected ion or atom beams.

E. Coincidence Techniques

Coincidence measurement techniques are widely usedin AMO physics. Application of similar techniques tosurface studies offers several advantages. The use ofpulsed ion, atom, or photon probe beams coupled withcoincidence measurements of outgoing ions orelectrons, or among the various products of theexcitation, could significantly enhance our knowledgeof energy flow and time evolution in surfaceprocesses, including the electronic states and atomicconfigurations involved in desorption, decomposition,and electron emission. One challenge in suchmeasurements is that surfaces are copious sources ofsecondary electrons, ions, and atoms, making thedetection of true coincidences more difficult. Otherproblems are the difficulty in efficiently collectingions or electrons from surfaces and the ability ofsurfaces to reneutralize and recapture species.Electron-electron and electron-ion coincidencemeasurements are currently being studied by a fewresearch groups. Expansion of such efforts would

provide new insights into the details of dynamicsurface processes.

F. Surface Studies in High PressureEnvironments

Most surface studies have been carried out in ultrahighvacuum environments, with possible pretreatment ofthe surface at slightly higher pressures so as to carryout chemical reactions. The ability to carry outmeasurements at higher effective pressures of up toeven 10-3 torr would permit duplicating the conditionsfound in semiconductor device and magnetic discfabrication and provide data of more technologicalrelevance. Optical probes are most readily adapted tohigh pressure environments since photons can travelthrough dense gases under appropriate conditions. Anexample is probing second-harmonic generation withlasers, fluorescence, and Raman techniques. Particlebeams present a greater challenge since they areattenuated in dense gases. The use of particle beamscould be achieved by using ionic or atomic beams ofsuitable flux and/or surrounding the sample with adifferentially-pumped cell whose aperturesaccommodate the entrance of the excitation source andthe exit of the detected products. With suitabledesigns, it might be possible to work at pressures inthe 5-10 torr range, thus more nearly approaching theconditions of atmospheric and environmental reactionson surfaces.

G. Higher-Efficiency and Higher-SpeedDetectors

Developments in particle and photon detectors nowallow single particle detection under a wide range ofconditions. However, this is not yet possible in someimportant cases, such as very large molecules andclusters (>106 Daltons) in mass spectrometry, atomsand molecules sputtered from surfaces, and low energyatoms and molecules important in new processingapplications. The development of higher-speed andhigher-efficiency detectors for atoms and ions requiresan understanding of how the kinetic energy of atomicmotion is transferred to electronic excitations in solids


43 Panel C

at low impact velocities, resulting in electron emission.Even with more intense beams of both photons andions, many experiments are detection-rate limited. Forexample, the rate of acquisition of photoelectronspectra is currently limited by the capabilities ofmultichannel detectors. For ions or atoms, detection iscomplicated further by the low and strongly energydependent efficiency of the detectors. The standardparticle detectors employed in particle/surfaceinteraction experiments are the channeltron multiplierand the microchannel plate array. These have nearunity efficiency for detection of electrons andenergetic ions at 10 keV, but the efficiency for ions oratoms monotonically falls to zero as the velocity of theions decreases. Developments in detector technologywould enormously expand the capabilities of manybeam scattering techniques, permitting more sensitiveand/or higher-resolution studies in many surfacecharacterization and surface analysis studies. Possibleapproaches for enhancing particle detection involvelaser ionization of the neutral beam and/ordevelopment of new surface coatings with higherelectron emission coefficients. Fundamental researchon detector development, including how the kineticenergy of atomic/molecular motion is most efficientlytransferred to electronic excitations in solid surfaces atlow impact velocities, would thus be very beneficial.

V. Key Research Areas in This ReportThat are Under-Represented in theUnited States

• Basic studies of the interactions of low energy(0.1eV - 0.5 keV), well-defined neutral and ionbeams with surfaces is vastly under-representedin this country. Germany, Japan, France, and TheNetherlands are presently leading the ion beamfield. The area is replete with importantapplications.

• Basic studies of the effects of low energy (1 eV -1 keV) electron beams with surfaces are under-represented in this country. This area also hasimportant applications.

• The STM was developed in Europe and novelapplications of scanning probes are continuallybeing demonstrated, especially in Germany andSwitzerland.

• Both Japan and Germany have substantiallylarger commitments to femtosecond laser work

in chemistry, biology, and materials science.

VI. Summary List of Basic UnsolvedProblems Involving the Interactionsof Low Energy Particles (Electrons,Ions, Atoms, Molecules, Clusters)and Photons with Surfaces

1. Theoretical treatment of excited states atinterfaces.

2. Atomically resolved structure determination ofsurfaces and adsorbed species.

3. Quantum-state-specific control and atomically-and chemically-resolved spectroscopiccharacterization and manipulation of surfaces.

4. Time- and spatial-resolved spectroscopicmonitoring of dynamic bond breaking andformation processes at surfaces.

5. Dynamics of energy and charge transfer atsurfaces, i.e. the flow of energy and charge andthe motions of atoms in particle-surfaceinteractions.

VII. Summary List of Applications ofParticle and Photon Interactions atSurfaces

1. Nonequilibrium growth, nanoscale fabrication,and processing of materials.

2. New techniques involving high intensity,spatially guided neutral and ion beams formesoscopic pattern deposition.

3. Mechanisms of surface chemical reactions andcatalysis.

4. Environmental problems, including sensing andprocessing of biological molecules.

5. Futuristic materials based on atomic andmolecular manipulations.


Panel C 44

VIII. Conclusion Statement

This report has identified some important researchneeds, opportunities, and applications that can beachieved through the interaction of two groups ofscientists, i.e. atomic, molecular, and optical physicistsand surface scientists. The synergistic expertise andtechniques from both groups provide an unparalleledopportunity to make significant advances in our basicunderstanding and application of the interactions ofenergetic particles (electrons, ions, atoms, molecules,and clusters) and photons with surfaces.

Panel D: Theory of Structure and Dynamics

Chair:Chris Greene (University of Colorado)

Panelists:Alexander Delgarno (Harvard Smithsonian Center for Astrophysics)

Charlotte Fischer (Vanderbilt University)Kate Kirby (Harvard College Observatory)

Kenneth Kulander (Lawrence Livermore National Laboratory)Uzi Landman (Georgia Institute of Technology)

James McGuire (Institut fur Kernphysik)Vincent McKoy (California Institute of Technology)

Anthony Starace (University of Nebraska)Malcolm Stocks (Oak Ridge National Laboratory)

David Yarkony (Johns Hopkins University)


Panel D 46

Executive Summary

In AMO physics, the terms “structure and dynamics”encompass diverse topical areas that intellectuallychallenge and relate closely to the mission of theDepartment of Energy (DOE). In these workingpapers, we summarize our findings. Our deliberationshave identified key themes in the theory communitythat overlap and further the research goals of BasicEnergy Sciences at DOE, broadly-construed. Thispanel has assessed new areas ripe for futuredevelopment within the next 5-10 years. Because ourcharge has been specifically aimed at identifying newresearch directions, we stress at the outset that manyexciting theoretical (and experimental) areas will notbe explicitly discussed. Other theoretical areas withgreat potential and relevance to the DOE mission arenot covered in this report. We have no room, forexample, to detail the varied and deep connections ofAMO science to the disciplines of astrophysics,atmospheric science, space science, environmentalscience, nuclear physics, particle and fundamentalphysics and the physics of reactive flows and shockwaves.

The areas of greatest excitement and potential for near-term development in structure and dynamics aregrouped into five categories: Control; AMO physicsin extreme conditions; Novel states of matter and light;Extending the frontiers; Opportunities and horizons inhigh-performance computing. We begin with a shortsummary of each of these categories. The rest of ourreport elaborates on each in somewhat greater detail.A continuing theme in all of these areas is a naturalprogression of knowledge and understanding fromsimple prototype atomic species to more complexspecies like molecules, clusters, and surfaces.

“Control” is an important theme for the future. Itembodies a clear long-range goal for AMO science.The field of coherent control, for instance, involves themanipulation of atomic and molecular processes by thedesign of femtosecond light pulses that cause thesystem to respond in a prespecified manner. Just as

the development of femtosecond lasers has broughtcoherent control to the brink of reality, so has theenhanced ability to manipulate atoms on surfacesbrought the potential of engineered nanostructures tothe forefront of this field. The techniques of atominterferometry and lithography will certainly play keyroles in the design of nanostructures.

“AMO physics in extreme conditions” includesunusual environments where basic scientific questionsremain inadequately understood. Atomic andmolecular collisions at ultracold temperatures, forinstance, probe the pure quantum limits. Ultra-intenselaser fields produce dramatic new physical effects andprocesses, while ultra-short light pulses are importantfor the “Control” goals discussed above. Newgeneration synchrotron light sources probe atomic andmolecular structure at ultra-high resolution (and insome cases at ultra-high energies), which magnifiesthe limitations of conventional independent-particlemodels and points to the need for improved theoreticalcapabilities.

“Novel states of matter and light” have energizedmuch of the scientific community in a fast-pacedexploration of striking new directions, largely drivenby recent experimental progress. Several groups cannow create BEC in their laboratories on a daily basis,but at the same time many basic theoretical issuesrequire a far deeper understanding. Quantum opticstheory and experiment, especially in the area of lasercooling of atoms, continue to show rapid, excitingprogress.

“Extending the frontiers” groups together severalareas that touch on or encompass current work in theBES AMO Physics program. New directions andcurrent bottlenecks deserving of priority effort arementioned, including the direct description of real-time dynamics for problems like intense laser-atominteractions, where the interaction is strong andlocalized in time, and must be treatednonperturbatively. Another area ripe for extension isthe theoretical study of relativistic correlation effects,


47 Panel D

both in heavy atoms and in simple moleculescontaining heavy atoms. These tie in with the DOEmission, as discussed below. Short-time dynamicalstudies using periodic orbit theory and scaled-variablespectroscopy can be applied to a range of newproblems. Finally, the interpretation of energeticplasmas requires a deeper understanding and thedevelopment of methods to explore a large number ofprocesses involving collisions of electrons, ions,atoms, and molecules.

“Opportunities and horizons in high-performancecomputing” includes areas where progress in AMOscience hinges on the utilization of large-memory,scalable multicomputers. The high performance ofthese powerful modern computers can dramaticallyenhance our ability to achieve robust descriptions offundamental processes and complex systems across awide spectrum of AMO science. Development ofscalable and paradigm-changing algorithms requiredfor efficient use of these new computing architectures,coupled with ready access to large parallel computers,can lead to major breakthroughs in computationallyintensive applications and can enhance the impact oftheory on experiment.

Elaboration on the Five Themes DescribedBriefly Above

Before we discuss the scientific directions andconnections identified by our panel, we believe it isimportant to recognize one of the constraints we facedin our undertaking. The separation of “Theory” intoone panel at this workshop, split off from the othertopical areas represented in this report, is unfortunatebecause the most productive advances in science—especially in AMO science—result from fruitfulsynergism between experiment and theory.Consequently, most of the forefront areas mentioned inthis section overlap closely with the areas discussed bythe other panels. In cases where the other panelspropose an area ripe for further development in thecoming decade, it should be understood that extensivetheoretical contributions can be very valuable forguiding and interpreting experiments. However,

theory does differ from experiment in “equipmentneeds.” Access to large-scale computational hardwareis needed by theorists in order to undertake certaintypes of cutting-edge calculations.

CONTROL

Coherent control

The dynamics of quantum systems can be modified bycoherent light pulses, and energy can be moved aroundin a predetermined way on an atomic scale. Advancesin short pulse laser technology have made it likely thatthe long-sought goal of quantum control of molecularcollision dynamics will finally be achieved. Inprinciple, manipulation of a time-dependent pulse willpermit the subsequent evolution of the system to becontrolled. Amplified few-fs Ti:sapphire pulses havefocused intensities larger than 1014 W/cm2, which isstrong enough to drive molecular transitionsefficiently. Selective dissociation, control of reactions,and preparation of desired molecular states orstructures can be achieved by these methods.Accomplishing these goals requires intimateinteraction between experiment and theory.Knowledge of the molecular potential energy surfaces,dipole coupling strengths as functions of moleculargeometry and the ability to determine the variations ofthe time evolution of the coherent wave packet causedby changes in the phases of the pulse will allowoptimization of the product yields. Precisedetermination of these required parameters exceedsexisting theoretical capabilities for all but the simplestdiatomic systems.

For more complicated systems, nevertheless, theorycan provide approximate starting points forexperiments designed to provide feedback to moreaccurately calibrate the system response, leadingiteratively to the desired results. One anticipatedexperimental problem is that the UV or VUVwavelengths required to induce an electronic transitionin most molecular systems are much shorter than thoseproduced by the Ti:sapphire amplifier. Harmonicconversion or frequency mixing techniques can


Panel D 48

generate coherent light pulses, as short as (or evenshorter than) the incident pulse, with almost anycentral frequency. Coherent control provides thecapability to probe and understand, at the mostfundamental level, the making and breaking ofchemical bonds in addition to enabling the productionof novel product states of the quantum system. Thiscan mean the generation of new chemical compounds,the establishment of more efficient routes to desiredreaction products or even the construction of particularstructures of the product molecules. This technologyhas applications beyond the modification of molecularcollision dynamics, such as the generation of Rydbergwave packets in atoms and the control of current flowin semiconductors.

Nanoscale Materials: Clusters, Nanocrystals,and Dots

Investigations of the microscopic physical originsunderlying size-evolutionary patterns of materialsproperties are among the main themes in moderncondensed matter physics and materials science.These studies are enabled by the emergence of novelexperimental probes and by the development of newtheoretical and algorithmic approaches. Coupled withthe availability of high-powered computational tools,they allow deep insights into the structure of matterbridging the atomic, molecular, and condensed-matterregimes. In addition to their basic scientificsignificance, the above issues form the scientific basefor technological developments, particularly in thearea of device miniaturization.

Atomic material aggregates with nanometer-scaledimensions exhibit certain properties analogous tothose found in other “zero-dimensional” (0D) systems,namely atoms and nuclei. Most prominent amongthese is the electronic shell structure of small(metallic) clusters. At shell closures this leads toenhanced stability and self-selection of clusters having“magic numbers” of atoms. Moreover, otherproperties such as ionization potentials, electronaffinities, and single-atom detachment energies ofopen-shell clusters exhibit odd-even alternations (as

functions of the number of atoms). These originatefrom cluster shape deformations similar to the Jahn-Teller distortions familiar in molecular (and nuclear)physics. Such clusters possess collective excitationspectra (giant photoabsorption resonances), and theyfission upon multiple ionization.

The adaptation of AMO physics methodologies andtheir application to clusters and nanocrystals, inconjunction with computer-based (time-dependent)first-principles molecular dynamics simulations, couldcontribute significantly to elucidation of the above-mentioned phenomena. Work in this area also bears onproblems pertaining to: supershells and semiclassicaldescriptions (using periodic-orbit theory) of electronicspectra for larger clusters; multiple charging (shaperesonances), basis-set selection and correlation effects;single-particle and collective (plasma) excitations;atom (or ion)-cluster and intercluster collisiondynamics, product branching ratios, and collisionalenergy redistribution; cluster and nanocrystal growthmechanisms; cluster thermodynamics (structural andphase transformations); magnetic properties; dipole(and higher multipole) excess electron binding tomolecular clusters [e.g. (H2O)n , n>2] and solvationenergetics and dynamics; and size-dependentreactivity. These studies could lead to formulation ofatomic-scale principles to guide the controlled designand preparation of nano-scale materials withprescribed size, shape, property, and functionality.

Another class of novel 0D nanoscale structures, withclose analogies to atomic systems, consists of confined(2D) electron gases in the form of dots. These areoften referred to as “super-atoms”. Both basic scienceand technology will be advanced by development ofan understanding of the electronic level structure, ofthe excitation spectra (both single-particle andcollective resonances), and of the charging propertiesin such systems. The improvement of theories beyondthe mean-field approximation is essential, both underfield-free conditions and under the influence of anapplied magnetic or optical field.


49 Panel D

Atom Interferometry and Lithography

Atom optics has shown promise in recent years, bothfor the fundamental interest it has generated and for itstechnological promise. Atom interferometry hasprogressed beyond the demonstration of basic two-and multi-slit interference patterns used as textbookillustrations of quantum mechanics. It can now beused as a tool to measure atomic properties includingthe phases of quantum amplitudes in addition to theirmagnitudes. The future applications of theseinterferometric techniques are hard to envision fully atthis time, but they will certainly probe the quantummechanical ideas of entanglement, coherence, anddecoherence. Experimental work in this area wouldbenefit from theoretical studies, particularly whennoise and dissipation play an important role in theinterference structures.

As technologies continue to improve for handlingphase-coherent atomic beams of small dimension, sowill the ability to lay down atoms on surfaces inprescribed arrangements. The continued developmentof these novel lithographic techniques is likely to havefar-reaching application to microchip production andnanoscale fabrication.

AMO PHYSICS IN EXTREMECONDITIONS

The theoretical description of AMO phenomena underextreme conditions generally requires the developmentof appropriate (and often new) theoretical approaches.Advances in experimental technologies have permittedmeasurements covering a broad range of suchphenomena, thus providing the opportunity for theoryand experiment to progress together. Many of theseopportunities have significant applications to energy-related technologies. The following are some of theleading opportunities:

Ultra Cold Collisions

At sub-milliKelvin temperatures, atom-atom collisionsenter a new regime in which de Broglie wavelengths

are far larger than atomic dimensions and in whichcollision times are longer than atomic radiative decaytimes. Atomic collision cross sections are dependenton long-range atom-atom interaction potentials. At thelowest energies, hyperfine interaction effects must beincluded. Acquisition of a deeper understanding ofthese cold collisions is crucial to the control andapplication of BEC phenomena, to improvements inthe precision and stability of atomic clocks, and toenhancements in the sensitivity of experiments todetect small effects and to make precisionmeasurements.

Ultra Intense Fields

At laser intensities greater than about 1012 W/cm2,nonlinear laser-atom effects (e.g., above thresholdionization, high harmonic generation, etc.) becomeincreasingly prominent; above about 1019 W/cm2, theelectric field in the laser light—rather than the nuclearCoulomb field—controls electronic motion and theinduced electron velocity becomes relativistic; aboveabout 1029 W/cm2, the laser field is strong enough to“spark the vacuum,” i.e., to create electron-positronpairs. The theoretical description of such phenomenarequires a non-perturbative treatment of the laser-atominteraction and, often, a time-dependent solution of thequantum equations for the laser-induced electronicmotion. Prominent among possible applications is theuse of high harmonic generation to produce coherentsources of VUV and X-ray radiation.

Ultra Short Pulses

The development of laser pulses as short as 5 fs is nowpossible, and shorter pulse durations are on theimmediate horizon. Pulses this short permit excitationsof localized electronic or vibronic wavepackets inatoms and molecules. The theoretical description ofthe electronic dynamics requires a time-dependentapproach. Furthermore, additional coherent, shortlaser pulses permit one to control the evolution of thesystem. Analysis of the wavepacket motion in atomsor molecules has important applications. In Rydbergatoms, theory can explore the connections between


Panel D 50

classical and quantum behaviors. In molecules, a goalfor theory and experiment is to develop robustmethods for controlling the evolution of theprobability density along particular desired Born-Oppenheimer pathways.

Ultra High Resolution and Detection Sensitivity

Lasers and third generation synchrotron light sourcesare permitting atomic and molecular spectra to beobserved with unprecedented resolution andsensitivity. Energy resolutions of 1 meV and better arebecoming available in the 100 eV photon energyrange; while this does not translate into a “resolution”figure as impressive as can be achieved with state-of-the-art CW laser systems, it is a major step forward inthe high energy regime. Cross sections as small ashundredths of a barn are being observed. In molecules,vibrational and even rotational energy levels can nowbe resolved even in some synchrotron experiments. Inatoms and in atomic negative ions, such high energyresolution and sensitivity requires theory to transcendthe independent-electron model to incorporate many-body correlation effects, as well as the effects ofnormally feeble interactions (e.g., polarization effects,spin-orbit and other relativistic effects). In molecules,theory at an elementary level has not been widelyapplied, even for relatively small systems.Photodetachment of negative ions leads to spectradominated by correlated two-electron resonances, withspectral patterns very different from the one-electronRydberg series omnipresent in neutrals. Knowledgeabout such resonance features is essential for a numberof energy technologies, such as in the design of gaslasers, arcs, fluorescent lamps, and high-powerdischarge switches, because resonances affect theconductivity of a low-temperature discharge andplasma environment.

New experimental techniques hold promise also incollision physics as they allow more detailed analysisof chemical kinematics and reaction dynamics. Onesuch technique is cold target recoil ion momentumspectroscopy (COLTRIMS). Analyses of dataobtained with this method have used classical

methods, but analyses using detailed quantumcalculations are largely incomplete. This is true notonly for many-electron systems but also for simplesystems at low velocities. A difficult challenge forquantum theory is to develop methods that go beyondthe independent particle model for reactions involvingmultiple electron transitions. Often these many-electron transitions are dominated by the dynamics ofelectron correlation, which is a key to understandingmechanisms for energy transfer in complex atomicsystems. The general question is how to understandcomplex systems (including large molecules ornanostructures) in terms of simpler atomic systems.

Ultra Fast Collisions

An example of a highly localized conversion of energyoccurs in interactions of relativistic ions with atoms inwhich electron-positron pairs are produced.Calculations have not yet been able to give a reliableinterpretation of the spectra observed, in which it isbelieved that disruption of the negative energy sea hasoccurred.

Ultra High Photon Energies

Another window of opportunity is being opened bynew third and fourth generation synchrotron radiationfacilities. These facilities are just beginning toproduce data useful in both the gaseous and condensedphases of matter. For example, Raman emission by x-ray scattering is being used to determine compositionof materials used in new semi-conducting devices.Interpretation of these experiments requires a theoristto consider both atomic and condensed-matter issues.The development of a microscopic understanding ofthe effects of high energy photons that penetrate matterrequires us to unravel multiple ionization andexcitation events produced by x-rays, both inphotoionization and in Compton scattering. Crosssections and reaction rates for these processes aredominated by the concerted motions of non-independent electrons, and photons probe the electroncorrelation dynamics cleanly. Sorting out suchprocesses could help us to understand how to modify


51 Panel D

or control site-specific reactions in more complexchemical and biological systems.

NOVEL STATES OF MATTER ANDLIGHT

Bose-Einstein Condensation (BEC) and atomlasers

Interest in AMO science intensified in 1995 followingthe first experimental observation of BEC for a dilutegas of spin-polarized alkali-metal atoms. The ensuingexcitement has had few parallels in the physicalsciences. Experimental laboratories worldwidequickly initiated their own attempts to reproduce thefirst experiments, and at last count six independentgroups had achieved BEC. While the initialobservation was primarily an experimentalachievement, theoretical studies have since played akey role in the continued growth and excitement of thefield. BEC theory is now a high profile activityrepresented in some of the best physics departments,and is carried out by both condensed matter and AMOtheorists. This subject has already spawned extensiveinterdisciplinary interactions that extend ourunderstanding of the many-body problem from theperspectives of both fields. Fundamental questionsarise relating to: the mechanisms of spontaneoussymmetry breaking and the formation of coherence;phenomena such as superfluid flow and persistentcurrents; specific heat and critical temperatures; firstand second sound waves and excitations; the kineticand dynamic evolution of the condensate.

This panel believes that theoretical efforts tounderstand BEC will help to unravel a number ofissues that need to be better understood such as: theenergy spectrum of this many-body condensate; itsformation and decay mechanisms; its coherenceproperties; the variety of novel forms in which it canbe created. A key question concerning its possiblerelevance to DOE interests arises from the simpleenergetics of BEC. What is usually termed the “groundstate” of the condensate is not actually the true groundstate of this many-body system; it is a metastable state

whose total energy is several electron volts per atomhigher than the ground state. In other words, thecondensate, despite being ultracold (50nK)translationally, stores a surprisingly high density ofinternal energy since about 106 atoms are confined to avolume of order 10-18 m3. Another type of proposedcondensate consists of helium atoms in metastableelectronic states; if experiments succeed in formingsuch a condensate, the amount of energy stored peratom would be increased by nearly an additional orderof magnitude. Should future research demonstrateways to achieve the controlled release of this energy,e.g. through an “output transducer,” this could be animportant source of coherent atoms. Such a phase-coherent beam of atoms was demonstrated earlier thisyear at MIT, and has been called an “atom laser.” Theatom laser beam is qualitatively different from anordinary incoherent atomic beam, and is likely tospawn spectacular capabilities for lithography andatom interferometry in the coming decade.·

Quantum Optics

Entanglement: Quantum optics brings fundamentalquantum mechanical theory to bear on thoughtexperiments that have recently been realizedexperimentally with the advent of high-precisioncavities and state-specific control of atomic statesthrough stable, tunable lasers. Micromaser cavitieswith exceptionally high Q-factors provide photontraps, analogous to particle traps, which are used tolocalize injected photon fields. These fields aretypically tunable over a wide range of frequencies nearresonance between two atomic levels of interest, e.g.,alkali Rydberg levels. The passage of one or moreatoms through such a cavity leads to an entanglementof the atom with the cavity, which can then be probedby a second atom, which in turn becomes entangledwith the first atom. Alternatively, two cavities can beentangled with one another via the passage of a singleatom, which might transfer a photon from one cavityto the other. All of these coherent processes aremanifestly quantum mechanical in nature, as isrevealed dramatically when the states are correlated


Panel D 52

with macroscopic detectors. Note that the use ofexplicitly nonclassical light fields (i.e. squeezed states)for high-precision frequency metrology is presentlyfunded in the DOE Atomic Physics Program.Related fertile areas include:

Dissipation theory: A central theme in quantum opticshas been the development of powerful theoreticalmethods that can efficiently handle irreversibility innonrelativistic quantum mechanics. These includequantum Monte Carlo and quantum trajectory methodsin which the continuous evolution of the Schroedingerequation is disrupted by quantum jumps of the state.The concepts are general and diversely applicable inquantum mechanics. In many-state simulations theyare far more efficient than the more standard “masterequation” techniques.

Quantum computing: When the usual binary logic ofcomputation is modified by quantum mechanicallyentangling two states, this leads to exponential gains incomputation speed for certain important problems. Theexperimental difficulties arising from decoherence forrealizing a quantum computer are formidable, but thepotential gains are extremely high. Proposals forquantum error correction, transfer and cloning ofquantum states, quantum communication, andquantum cryptography are all interesting by-productsof this research. It has been shown that arbitraryquantum states can be engineered, and using recently-developed tomographic methods, experiments cancharacterize these states completely.

Laser cooling and atom optics: The interaction ofatoms with the electromagnetic field allows trappingand cooling and to create optical elements such asbeam splitters and mirrors for matter-waveinterferometers. There have been tremendousexperimental advances in laser cooling techniquesover recent years with the lowest temperatures givenby dark state cooling in which the coldest atomsbecome decoupled from the field through a destructiveinternal interference of the absorption amplitude.Atoms cooled in this way undergo Levy flights and

have interesting and anomalous statistical properties.Experiments on lattices of light that confine atoms in athree-dimensional array of optical potential wells areimportant for precision measurements.

Antimatter and exotic species

Experimentalists and theorists have recently studiedthe spectroscopy of antiprotonic helium. Effortscontinue toward creating antihydrogen at low energiesfor tests of fundamental physics. A dedicatedantiproton storage ring is being built at CERN, whichwill enable the collection of antiprotons in a Penningtrap. As these experimental capabilities leap forward,the branch of atomic physics concerned with thestructure and collisions of exotic particles at lowenergies has many opportunities and challenges, inwhich theoretical contributions are essential.Enormous amounts of energy could be stored in theform of antimatter, if its creation and manipulationadvances tremendously beyond present-daycapabilities, though this appears too remote to serve asa goal for the immediate future.

EXTENDING THE FRONTIERS

Time-Dependent Real-Time Dynamics

The dynamics of quantum systems can often bedescribed directly in the time domain. This approachprovides accurate product-state distributions over a rangeof total energies from a single calculation. When coupledwith modern visualization tools, it can provide keyinsights into the underlying mechanisms. For manyprocesses this is the only practical formulation, especiallyin cases for which the number of open channels is veryhigh (e.g. when three-body continua are present) orwhere the Hamiltonian for the system is explicitly time-dependent (e.g. when an intense laser pulse is applied).For even the simplest systems this approach can requirevery substantial computational resources. Utilization oflarge multi-processor computers has allowed theinvestigation of many phenomena for simple systemsthat are central to the missions of the DOE. Many of thecurrently-used methods solve the time-dependentSchroedinger equation iteratively—a scheme that is


53 Panel D

ideally suited to parallel architectures. The complexity ofproblems that can be addressed in the future is contingentupon the development of efficient algorithms thatmaximize the division of labor among numerous fastprocessors. Problems that have been successfullyinvestigated include molecular dissociation by electron,ion, or photon impact; inelastic and reactive atomic andmolecular scattering processes; atomic excitation andionization by intense, pulsed laser fields.

An outstanding problem in this area is the study ofmultielectron systems subject to a time-dependentinteraction: either a collision or a pulsed electromagneticfield. Even for two electron systems, there are manyproblems that have not yet been solved. Theindependent-particle approximation often failsdramatically in a time-dependent problem, and is thusinadequate. Resources and numerical methods arepresently being developed that should eventually permitan accurate numerical solution of the full-dimensionaltwo-electron problem for a number of non-trivialregimes. Among these is the surprisingly strongsimultaneous double ionization of helium by a shortpulse optical laser field: interpretation of theseexperiments continues to resist theoretical efforts. Thecalculations challenge our ability to solve the time-dependent Schroedinger equation, of course, but perhapsof equal importance, they challenge our visualizationskills needed to identify the dynamical mechanisms thatoperate. A desired outcome is that, based on the insightgleaned from a solution of the dynamics for two-electrons in a laser field, a general improved techniquewill be developed that can handle similar nonperturbativeproblems for multielectron systems.

Relativity and correlations in structure anddynamics

Two active research areas that have been traditionallyaddressed within current AMO theory projects in thisprogram include electron correlations and relativisticeffects. The study of correlations aims broadly at thederivation of theoretical tools capable of describingnumerous atomic phenomena that are beyond thescope of the independent-particle approximation: the

binding of an electron to heavy atoms with closedsubshells; the escape of two or more electronsfollowing absorption of a single photon; collectiveexcitations of two-electron resonance states.Numerous successes of this line of research can beidentified. Theoretical capabilities have advancedtremendously, especially in the last few years. Whilethis field maintains a continuing interest in furtherimproving theoretical methods for even the simplestatoms with only two or three valence electrons, wefocus here on new advances that are increasinglydesirable. This is truly a frontier, a theoretical subfieldthat remains, comparatively speaking, in its infancy:the description of electron correlation effects in boundand continous spectra where relativistic effects areimportant. This includes mainly the atoms heavierthan, say, Kr (Z=36), as well as diatomic and triatomicmolecules containing one or more of these heavyatoms.

In the atomic realm, the most commonly-usedapproaches start with the Dirac-Coulomb Hamiltonian,and a multiconfiguration Dirac-Fock wave functionexpansion which, in the case of fixed basis setmethods can be obtained through configurationinteraction (CI) calculations. Further effects can beadded perturbatively. Despite the ability to treataspects of the low-lying bound state spectrum for farmore complex open-shell atoms and negative ions thanwas conceivable ten years ago, roadblocks loom. Tobe sure, the spectra of these species are extremelycomplicated, especially in the lanthanides (currentlyimportant for the lighting industry) and in the actinideand transuranic elements (of continuing relevance tothe DOE environmental cleanup mission). Forexample, in Lr (Z=103), theoretical investigationshave concentrated on the determination of the groundstate properties, but even the ionization energy is notknown accurately. The number of configurations thatarise, even in the most constrained CI expansions,grows explosively in these systems with one or moreopen subshells. In fact, it is worth asking whethercalculations based on this methodology truly have arealistic hope of being able to calculate accurate


Panel D 54

energy levels, photoabsorption and photoionizationoscillator strengths, dielectronic recombination, andhyperfine and weak-interaction couplings, at a level ofprecision that is spectroscopically useful. Multichannellevel perturbations are so rampant that it may makesense to pursue the more limited initial goal ofdescribing short-time spectra, i.e. the spectroscopy at acomparatively coarse-grained energy resolution.

Many of these same bottlenecks arise in the theoreticaldescription of a diatomic or triatomic molecule, ofcourse. Additional complexities arise because of themuch greater difficulty of performing an electroniccalculation in a multicenter geometry, and with theintimidating complexity of the nuclear dynamics onnonadiabatically-coupled potential surfaces. Intriatomics, conical intersections lead to all thedifficulties, e.g., of Jahn-Teller effects that modify themolecular symmetry one would expect based on fixed-nuclei calculations. Atom-diatom scatteringcalculations can sometimes describe these effects withmodest computational effort, and in a conceptually-appealing manner, by using geometric phasesappropriately. Moreover, while the construction ofbound-electronic potential surfaces and a descriptionof the nuclear dynamics is a tremendous challenge,one would like the theory to advance and develop thecapability to treat electronic continua that arise inphotoionization, photodetachment, and electronscattering processes. In the molecular realm, muchremains to be understood even for moleculescontaining the first-row transition metal atoms, whererelativity can still be handled perturbatively.

The research in this area requires extensive conceptualand computational development. We believe that it ispoised to make tremendous strides in the comingdecade, by coupling the deeper understanding we haveachieved in how to treat these types of bound andcontinuum problems with the rapid improvements incomputational hardware and algorithms.

Quantum Mechanical Approximations Derivedfrom Semiclassical Methods

A minor revolution has occurred during the pastdecade in the theoretical description of quantumsystems with two or more strongly-coupled degrees offreedom. The most exciting possibilities arise insystems whose classical Newtonian analogues arechaotic. Classical chaos arises when classicaltrajectories are exponentially sensitive to initialconditions. The quantum dynamics for such systemsis sometimes referred to as “quantum chaos.”

Our panel has for the most part avoided a discussion oftheoretical possibilities based on the merits of anyparticular “technique.” The semiclassical “quantumchaos” methods devised by Gutzwiller, Berry, andtheir followers seem to us worth singling out, however,because their new capabilities complement moreconventional schemes used by AMO theorists. Inparticular, new types of spectroscopies (such as scaled-variable spectroscopy) give deep insight into the originof short-time features in the “time domain” of theobservables such as photoabsorption cross sections.Of course, short-time features translate via Fourieranalysis into global, coarse-energy features; this isoften of greatest qualitative interest when anexperimentalist seeks an interpretation of a hopelesslydense spectrum of innumerable lines. Quantum chaostechniques interpret the main global features in termsof a modest number of periodic classical orbits. Thenew capabilities of such theories are deserving ofspecial scrutiny from this scientific community. Theirfull domain of applicability remains to be identified. Itis encouraging that for a class of problems involvingRydberg state dynamics in external fields, they haveprovided unparalleled insights into the origin of theshort-time spectrum.

For decades, applied mathematics has been exploringproblems relating to the asymptotic distribution ofeigenvalues, a classical area in mathematics that hasbeen reinvigorated by periodic orbit theory. Moregenerally, theoretical atomic physics would also


55 Panel D

benefit from stronger interaction with the largerapplied mathematics community. That communitycontains expertise in areas of key relevance to atomicphysics, including differential equations, numericalmethods, and asymptotic analysis.

Low-Temperature Plasma-Related Research

The problems presented in understanding the physicsof high and low temperature plasmas lie at theforefront of atomic, molecular and optical physics.The behavior of plasmas is determined by a complexof interactions of a wide array of atomic, molecularand optical processes involving electrons, ions, atomsand molecules. The full characterization of many ofthe processes that determine the nature of plasmas waitupon advances in theoretical concepts anddevelopments in computational methods. Thermalfusion plasmas, inertial confinement plasmas andplasmas designed to simulate weapons effects are hot,with temperatures approaching one million to tenmillion degrees. Similar hot plasmas occur naturallyin the solar corona, in supernova remnants and in gassurrounding cosmic X-ray sources. The behavior ofhigh temperature plasmas is determined by ionization,recombination and excitation processes. Theefficiencies of many of the processes can be calculatedby the application of methods that have been alreadydeveloped and tested. For others, and particularly forsystems of high nuclear charges, further developmentis needed that takes fully into account relativisticeffects. Using the data, computer-based models thatsimulate the behavior of the plasmas can beconstructed. The plasma properties can then beexplored over a wide range of conditions of density,temperature and radiation environments and diagnosticprobes can be designed.

Lower temperature plasmas are created formanufacturing purposes in the lighting and plasmaprocessing industries. They occur naturally in theionospheres of the planets, in lightning discharges inthe terrestrial atmosphere and in the interstellarmedium. The behavior of cooler plasmas withtemperatures below about 5000 K is complicated by

the presence of neutral and ionic molecules andnegative ions. Dust may be formed and will exercise amajor influence on the characteristics of the plasmas.The molecular composition is drastically modified bythe production of metastable species and by thevibrational excitation of the molecular species.Important processes are dissociative recombination,dissociative attachment, associative detachment,collisional dissociation, and chemical reactionsinvolving excited species. Few data are available andfor many of the processes no adequate theory has beenformulated. The processes of dissociativerecombination (DR) of molecular ions with electronsand charge transfer reactions of ions with molecules,in particular, pose formidable theoretical challenges.For even some of the simplest diatomics (HeH+) andtriatomics (H3+), theory and experiment disagree aboutthe near-threshold DR cross sections by more than anorder of magnitude.

Present models of plasmas contain only a limiteddescription of the processes and are based onunreliable data. With anticipated advances in collisiontheory for complex systems and access to enhancedcomputing capabilities, it will be possible to identifythe critical processes, to calculate their efficiencies andto construct models that can be used to guide thedesign of low temperature plasmas for specificapplications.

Ion-Atom and Ion-Molecule Collisions

New data are becoming available for many differentkinds of reactions. As the possibilities of observingtransitions and reactions in many-electron systemsgrow, so do the difficulties of finding which studieswill provide the best insight into the nature ofdynamical processes in complex systems. In thesecases theory can often provide guidance in selectingreaction details most crucial to understandingmechanisms for dynamics in complex systems. Achallenge in the field of collisions is to unravelreaction mechanisms among many-electron atoms,molecules, clusters, and surfaces, including instancesin which perturbation techniques fail. In this field


Panel D 56

experiments carried out at ion storage rings haveprovided critical data at unprecedented resolution for anumber of electron-ion recombination and collisionprocesses. This is an area in which the United Stateshas fallen significantly behind development in Europe.Smaller, less expensive, “table-top” storage rings,which should be available soon, provide an importantopportunity for U.S. science to explore many of thesetopics.

OPPORTUNITIES AND HORIZONS INHIGH-PERFORMANCE COMPUTING

Progress in this century in atomic, molecular andoptical theory has placed the field in an enviableposition to make accurate predictions aboutfundamental processes. However, for the complexsystems and processes of increasing relevance to thisfield, the impact of theory on experiment can beincreased dramatically through use of the mostpowerful modern computers and computingtechniques. The significant increase in performanceprovided by such computers will enable AMO theoryto address fundamental processes in more complexsystems. Furthermore, as many branches of AMOtheory make the transition to parallel computing,within a decade the applications developed for today’smassively parallel computers will be running onwidely-available laboratory-scale multiprocessors.

The accurate theoretical description of collisionsamong electrons, ions, atoms, molecules, clusters andsurfaces cannot be achieved without this scale ofcomputation. Such advances will allow us to predictthe rates of the many important collision processes andto construct simulations of the low-temperatureplasmas used widely in processing materials. They willalso permit studies of bound and continuum states ofheavy elements and of new phenomena occurring incomplex systems in intense fields. A further impact ofhigh-performance computing is the opportunity tomake the computational tools of modern theory moreaccessible to the experimental community by movingbeyond the “grand challenge” stage to the point where

use of such tools pervades the entire AMO community.Many AMO codes (RMATRIX, MCHF, GRASP, andCowan’s suite of programs) have been developed andused by theory groups, but little attention has beengiven to making them user-friendly and readilyaccessible to experimentalists. Other communitiessuch as the chemistry community have begun to makethis transition. Here, quantum chemistry codes such asGAUSSIAN, GAMES, QCHEM and others have beenbecome the every day tools of organic and inorganicchemistry. Few similarly-accessible tools exist for theinterpretation of experimental results for a wide rangeof problems that involve dynamics, such as electroncollisions, photoionization, and molecular and ioniccollisions. With the advent of new DOE nationalexperimental facilities such as the Advanced LightSource and the Advanced Photon Source, the need forsuch tools and for closing the gap between theory andexperiment is more urgent than ever.

Panel E: Nano- and Mesoscopic Structures

Chair:Paul Alivisatos (Lawrence Berkeley National Laboratory)

Panelists:Daniel Chemla (Lawrence Berkeley National Laboratory)

Vicki Colvin (Rice University)Robert Compton (University of Tennessee)

Therese Cotton (Iowa State University)James Heath (University of California, Los Angeles)Mark Knickelbein (Argonne National Laboratory)

Charles Lieber (Harvard University)Paul McEuen (University of California, Berkeley)Horst Schmidt-Boecking (Universitat Frankfurt)

Stanley Williams (Hewlett-Packard Laboratories)


Panel E 58

Solids of nanometer scale exhibit strongly sizedependent chemical and physical properties whichrepresent limiting behaviors for different types ofmatter (atomic to bulk). The variations with size areenormous, and this represents a new opportunity tooptimize material properties by varying their size andshape rather than by changing their chemicalcomposition. The developments in nanoscale scienceare of a sufficiently fundamental nature that we cananticipate the development of this field willsignificantly impact several other disciplines of basicscience and will also help address key energy relatedtechnologies.

Advantages and Challenges of Nanometer ScaleScience and Technology

The ability to shape and control matter on thenanometer scale offers a tremendous number ofestablished and potential advantages, but to fullyexploit this area there are a number of critical issuesthat must be addressed over the long term. In thissection of the report, we discuss both the majoradvantages that nano-scale based systems offer to newand/or improved technologies, as well as the majorquestions that surround the development of any suchnew technologies.

The primary advantage of any nanostructured materiallies in the extensive tunability of its properties. Forexample, the physical and chemical characteristics ofmaterials have traditionally been determined by thechemical stoichiometry and structure of the material.However, by controlling the size and shape of a nano-scale solid, many of the physical and chemicalproperties of the system can be tuned over a largerange. For example, the fundamental characteristics ofa material, such as its melting temperature, color,saturation magnetization and coercivity, chargingenergy, chemical reactivity, etc., are all a function ofsize and shape. For instance, the color ofsemiconductor quantum dots can be variedcontinuously from the near infrared to the ultraviolet.Such color changes correlate to electron and holeenergy levels, which in turn affect the catalytic and

chemical behavior of the particles. Thus, nano-scalebuilding blocks lend major new experimentallycontrollable variables for fabricating desired materials.In the last decade, it has been possible to fabricatenanostructured materials at a level where they nowcompare in quality to the high performance solids andthin films used in the electronics industry. A partiallist includes highly monodisperse colloidal metal andsemiconductor nanocrystals, STM-assembled“quantum corrals,” molecular beam epitaxy grownself-assembled quantum dots, laser ablation and clusterbeam expansions, electric discharge growth of carbontubes and clusters, thermal flow reactors, rapidpyrolysis, as well of course as e-beam and x-raylithography. Atom manipulation, and matterdiffraction from light waves, are important new toolsemerging from atomic and optical physics which maylead to new ways of fabricating nanostructures.

Despite the wide range of fabrication methods, a highdegree of structural perfection is often observed innanostructures, and this can be attributed to severalfactors: In a bulk single crystal, it is impossible toremove all defects. In a nanostructure, the highsurface area provides a route for the annealing andremoval of defective sites on short time scales. Relatedto this is the fact that the melting temperature of solidsis substantially reduced in finite size, which allows notonly for the fast diffusion of defects, but also forgrowth schemes at surprisingly low temperatures.Thermodynamically, the probability of defectformation dictates the fraction of atoms likely to be indefect sites; when the total number of atoms in thesystem is small, this number approaches zero. Thesearguments suggest that the manufacture ofnanostructures can be done very cheaply, with simpleprocessing. This makes nanostructures extremelyattractive candidates for a host of technologies.

As the quality and variety of nanostructures hasimproved over the last decade, many newtechnological applications have been envisioned. This,in turn, has raised several questions related to theintegration and performance of nanostructures in


59 Panel E

various technologies. First, can sufficient quantities ofhigh uniformity nanostructures be made to enabletechnological applications? All current fabricationschemes operate on a typical laboratory scale (rangingfrom 1 monolayer to 1 gram). Second, one of theadvantages of nanostructures is that they arecharacterized by a very high surface area. However,this also makes interface characterization and controlimperative. Opportunities for the development of newspectroscopic and imaging tools exist here. There is asyet no tool available to determine the bond lengths,angles and reconstructions of nanoparticles, as existsfor plane single crystal surfaces. Third, manyenvisioned applications require nanostructures to beplaced into chemically and/or structurally complexenvironments – i.e. around other nanostructures,connected to electrical devices, in chemical sensingenvironments, etc. The assembly of nanostructuresinto specific architectures is a challenge that is justnow beginning to be addressed. Fourth, the chemicaland thermodynamic stability requirements fornanostructures will vary from application toapplication. This issue remains largely unexplored.

Nanostructures are extremely promising newmaterials, with significance for both basic and appliedscience. Although the field of nanostructures is young,it is already possible to identify key areas in whichnanostructure science must evolve for the field toreach its full potential. At present time, none of thequestions discussed here appear insurmountable.

Impact of Nanoscience on FundamentalProblems

Nanostructure studies are most often described interms of their potential technological impact, to thepoint where it is a field often referred to asnanotechnology. Yet the ability to control matter onthe nm length scale can be expected to have importantconsequences for fundamental studies in disciplinesranging from condensed matter theory, to solid statechemistry, to materials science.

An isolated nanometer scale object such as ananocrystal or a nanotube can be viewed as anartificial atom, possessing a well-defined charge stateand energy level structure. Unlike real atoms,however, the size, shape, properties, and localenvironment can be controlled and adjusted. As aresult, nanostructures are ideal model systems forexploring the physics of small quantum systems.

The electronic structure of the delocalized electrons ina nanostructure can be continuously tuned by adjustingthe size of the object. This shifts the absorption andemission spectrum from the bulk material and resultsin discrete lines. Further, the charge state of thenanostructure can be tuned using a voltage applied toan external electrode, continuously adjusting theionization energy of the artificial atom. Finally, theshape can also be adjusted. For example, 3D sphericalnanocrystals, 2D disc shaped quantum dots, and 1Drod-like molecules are currently under study toinvestigate the energy level structure of systems ofdifferent dimensionality. This tunability makespossible the systematic investigation of the relativeimportance of Coulomb interactions and quantumconfinement effects in small quantum systems. It alsomakes possible the creation of artificial quantumsystems with properties that are specially chosen for aspecific scientific application. This ability to controlthe electronic structure is of considerable interest fromthe point of view of interaction of light with matterand quantum optics. The non-linear optical propertiesand the question of electronic coherence in dot excitedstates are particularly related to developments inatomic and optical physics. For example, dots arepossible candidates as qu-bits in quantumcomputation.

The artificial atoms can also be assembled intoartificial molecules and solids, and the properties ofthis solid can be controlled. In particular, the levelstructure of the individual nanostructures, theircharging energy, their Coulombic and tunnel coupling,and their coordination can all be tuned. This istremendously important for understanding the


Panel E 60

behavior of interacting electrons in reduceddimensions, which is probably the central task facingcondensed matter physics. From high temperaturesuperconductors to Luttinger liquids, the non-Fermiliquid ground states that are believed to occur in 2Dand 1D interacting electron gases remain poorlyunderstood, both due to a lack of tunable experimentalsystems and their intrinsic theoretical difficulty. Theoptical, electrical, and magnetic properties of suchstructures will be of great interest.

Nanoscale materials provide many opportunities forimproved understanding of basic chemical processes.The study of chemical reactions occurring on andwithin nanoscopic objects such as clusters, dots andnanotubes provide a unique opportunity for examiningthe effects of reduced dimensionality. One examplewith environmental implications involves the study ofchemical reactions of atmospheric pollutants withinwater nanodroplets and on the surface of icenanocrystals. A second example concerns ourunderstanding of solid state chemistry. Solid statesynthesis is generally thermodynamic, largely becausemechanistic thinking is difficult to implement ininhomogeneous solids with defects present.Nanostructures provide the opportunity to prepare pre-assembled, multi-component single grains, in whichthe kinetics and mechanisms of solid state processescan be independently measured and controlled. Thestudies of metal cluster nanosurfaces provide anopportunity to examine intrinsic (electronic ofgeometric) size effects on surface reactions of catalyticimportance such as Fischer-Tropsch synthesis andhydrodesulfurization. The newly developed ability toput stable molecules as well as molecular coreactantsin finite cluster solvents, and in liquid heliumnanodroplets opens the possibility performingspectroscopic and reactivity studies at temperaturesbelow 1K.

Nanometer scale materials such as semiconductornanocrystals and carbon nanotubes can be created withno defects. This allows for fundamental materialsscience studies of the properties of pure materials, e.g.

the strength or structural metastability. In bulksystems, these properties can be influenced bydefects. The ability to create perfect materials makesmeasurements of the intrinsic, i.e. defect free,properties possible. The mechanical properties ofultrasmall structures are also of fundamental interest,e.g. size quantization of phonon modes and quantum“mechanical” oscillators. A new field ofnanomechanics is emerging that explores the physicsand technology of mechanical structures at this scale.

The Role of Theory

Theory plays a central role in the description of thestructure and dynamics of nanostructures as well as topredict the growth of such materials, and can help toguide experiment. In turn, nanostructures provide animportant proving ground for new theoreticaldevelopments. The ability to describe complexsystems is of particular theoretical interest, and theevolution with size of nanostructure propertiesprovides a heirarchy of complexity.

The theoretical description of isolated nanostructuresis just beginning. Vastly different computationalmethods such as ab initio quantum chemicalcalculations, density functional theory coupled withmolecular dynamics, and simple tight bindingapproximations all yield electronic ground and excitedstate energy levels for neutral and charged isolatednanostructures. In some cases the applications ofsemi-empirical methods combined with experimentalbenchmarks can be used to generate useful predictionsof a wide range of related structures such asfunctionalized nanoparticles. The issue of predictingequlibrium structures, including geometryoptimization and surface reconstruction including thepresence of ligands or other interface atoms isparticularly important. A number of theoreticalmethods can be applied to the mechanisms andkinetics of nanostructure growth. The calculation ofvibrational and electronic density of states as well aspolarizability and multipole moments can be comparedwith experiment. Of particular importance is theprediction of magnetic properties of nanostructures.


61 Panel E

The effects of geometry changes and external stress onthese properties is of interest.

The physical and chemical mechanisms wherebynanostructures, for example quantum dots, are coupledto the environment and to each other are of greatinterest. Properties such as electron transfer ratesbetween quantum dots, artificial solid band structuresand collective excitations in coupled dots need to beaddressed. There are many questions regarding theinteractions of dots coupled to the environment.Empirical potentials for dot-dot and molecule-dotinteractions together with the long range forcesresponsible for the assembly of dot arrays areimportance. Of equal importance is the chemicalreactions occurring at dot surfaces.

Relationship to Energy Technologies

Almost every basic energy related technology may beimpacted by developments in nanoscale science.These include: chemical technology, energy storageand manipulation, information storage andmanipulation, and structural and mechanical materials.

Catalysis inherently relies on nanoscale componentsand can be expected to benefit greatly from the abilityto rationally control the size and shape of componentswith precision and to spatially organize thecomponents. Nanostructures are also exquisitelysensitive to changes in their chemical and physicalenvironment, making them strong candidates forsensor elements; for this to be realized issues ofselectivity and amplification these must be addressed.Nanostructures also provide a model system forstudying reactions on aerosols, which in turn areimportant for understanding atmospheric chemistryand pollution remediation.

Photovoltaics, light-emitting diodes, batteries, fuelcells and capacitors are some of the energy storageand manipulation technologies that may benefit fromthe high surface to volume ratios provided bynanostructures. Nanostructures have the desirablephysical properties of solid systems with the added

benefit of the ease of processing of molecules.Composites of inorganic nanoparticles with organicpolymers are of particular interest in this category, andevery type of device mentioned above has beendemonstrated in composites

A bit is a physical entity, and there is an energy costassociated with information storage and manipulation.In the past 25 years the energy efficiency andconsequently the speed of computation per inputpower have increased by a factor of ten thousand.Silicon integrated circuit technology should be able todeliver another factor of one hundred increase before itreaches technological limits. However fundamentalphysical limits indicate that computers can improveanother factor of ten billion in energy efficiency. Toapproach the thermodynamic limit it will be necessaryto construct computers from nanostructures andseveral possible models for subsystems already exist.Computing will benefit tremendously from thecombination of decreased energy consumption,increased speed, and enormous storage increases (x1010) that will be enabled by nanostructures.

Magnetism is fundamentally different in nanosizegrains, which act as single domains. If the shape andarrangement of magnetic nanostructures can becontrolled, this will have uses in magnetic storage andsensors.

Nanoscale building blocks tend to spontaneouslyexclude defects. As a consequence, the structural andmechanical properties of the individual componentsand of their assemblies are quite distinct from those ofextended solids. The striking example of this is carbonnanotubes which appear to be the strongest fiber everdiscovered. Nanostructure composites with welldefined architectures will have unusual and desirablemechanical properties.

Relationship of AMO Physics to Nanoscience

Atomic, molecular and optical physics has contributedgreatly to the development of the emerging field ofnanoscience. Nanostructures and clusters, in fact,


Panel E 62

represent a “bridge” between the isolated atom ormolecule to the condensed matter state. Following thedevelopment of nozzle-jet expansions for theproduction of gas phase clusters, many AMO scientistshave contributed to our understanding of such diversephenomena as solvation, the hydrated electron,chemistry of nanosurfaces, etc. More recently, theassembly of nanostructure materials at themacroscopic level has produced a heightened level ofinterest in the AMO community. Many of thetraditional techniques inherent in AMO physics arebeing directly applied to the study of nanostructures,e.g. mass spectroscopy, electron spectroscopies,photon spectroscopy, etc. In fact, the discovery of C60

came about through the combination of a number ofthese techniques: (1) laser ablation of graphite into anozzle jet expansion, (2) laser multiphoton ionizationof C60 followed by (3) time-of-flight massspectroscopy. In addition to these standardtechniques, there are a number of new advances inAMOP techniques which might be important to thedevelopment of nanostructure science:

Many possibilities exist for transferring experimentaltechniques from atomic physics to nanoscience.Recent developments of laser cooling of atoms mightbe applied to the partial cooling of nanostructures.The ability to cool these clusters would lead to a betterinterpretation of their spectroscopy. It is alsoconceivable that such cooling could lead to newmolecular states of matter.

The study of the ionic properties of nanostructuresrepresents an exciting new field for AMO physics.Many of the ionization processes operative innanostructures do not occur on the atomic ormolecular level. For example, the ionization ofclusters such as C60 occurs through surface plasmonexcitations. The production of new “super atoms”such as metal endohedral fullerenes will afford studiesof energy levels of interest to solid-stateheterostructures. The attachment of electrons tonanostructures represents a new field of physics inwhich techniques of AMOP are essential.

Nanostructures provide templates for the first detailedstudies of multiply charged negative ions.

Many developments in optical physics may impactnanoscience. As one example, the “holy grail” of laserchemistry is to coherently control the products ofreactions by the application of shaped optical pulsesequences. This is a very active research area. It ispossible that advances in this field could contribute tothe “assembly” of nanostructured materials, or to themanipulation of electronic coherence in quantum dotmolecules and solids. The interaction of intenseoptical fields with nanostructures is of interest.Multiphoton ionization, and the measurement ofnonlinear susceptibilities as a function of the size arecurrently actively investigated. The ability to controlsize may provide more insight into the differencesbetween atoms and molecules in their non-linearresponses.

It is possible to reduce the size of electricallycontrolled macroscopic objects to sufficiently smalldimensions that the nanostructure can be manipulatedby the application of an external electrical field, e.g. avery high field applied between two neighboringquantum dots can directly influence the reactions ofthese quantum dots with their neighbors or scatteredprojectiles. In such strong fields, collisions of atomsor molecules with the nanostructures might allowelectron transfer from the quantum dot to the collidingatom or tunnel directly to vacuum states. Suchenhanced electron emission processes might findapplications in, for example, electron-inducedcatalysis, air pollution abatement, etc. By varying thesize, geometry, and external field one may also obtaindirect information regarding the energy levels ofnanostructures.

Traditional electron, ion, atom, etc. beams can be usedto excite or even fragment these nanostructures todetermine their structural and dynamical properties.In addition, new atomic beam microscopies (e.g. coldsupersonic He beams or atom laser beams) could beused to study the topography and phonon structure of


63 Panel E

the surface. Using seeded supersonic cluster beams, orpossibly laser cooling, very cold nanostructures maybe created.

Cold atomic beams can create interference patterns ofatoms with order 100 nanometer resolution. In thiscase, the AMOP techniques for producing such coldbeams are directly applicable to lithography.Furthermore, using UV or even low energy x-raysfrom synchrotron light sources, 3-dimensionalnanostructures or layers can be produced.

atomic, molecular, and optical physics workshopastro1.panet.utoledo.edu/~ljc/amopfr.pdf · atomic,...

Documents