supercomputing takes another giant step/67531/metadc741360/... · breakthrough in laser glass...

SupercomputingTakes AnotherGiant Step

SupercomputingTakes AnotherGiant Step

Also in this issue: • Secrets of Actinides• Predictable Aerogel Structure• Tibet—Where Continents Collide

June 2000

U.S. Department of Energy’s

Lawrence LivermoreNational Laboratory

About the Cover

About the Review

Lawrence Livermore National Laboratory is operated by the University of California for theDepartment of Energy. At Livermore, we focus science and technology on assuring our nation’s security.We also apply that expertise to solve other important national problems in energy, bioscience, and theenvironment. Science & Technology Review is published 10 times a year to communicate, to a broadaudience, the Laboratory’s scientific and technological accomplishments in fulfilling its primary missions.The publication’s goal is to help readers understand these accomplishments and appreciate their value tothe individual citizen, the nation, and the world.

Please address any correspondence (including name and address changes) to S&TR, Mail Stop L-664,Lawrence Livermore National Laboratory, P.O. Box 808, Livermore, California 94551, or telephone (925) 423-3432. Our e-mail address is [email protected]. S&TR is available on the World Wide Web atwww.llnl.gov/str/.

© 2000. The Regents of the University of California. All rights reserved. This document has been authored by theRegents of the University of California under Contract No. W-7405-Eng-48 with the U.S. Government. To requestpermission to use any material contained in this document, please submit your request in writing to the TechnicalInformation Department, Document Approval and Report Services, Lawrence Livermore National Laboratory, P.O. Box 808, Livermore, California 94551, or to our electronic mail address [email protected].

This document was prepared as an account of work sponsored by an agency of the United States Government. Neitherthe United States Government nor the University of California nor any of their employees makes any warranty,expressed or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of anyinformation, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights.Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, orotherwise, does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United StatesGovernment or the University of California. The views and opinions of authors expressed herein do not necessarily stateor reflect those of the United States Government or the University of California and shall not be used for advertising orproduct endorsement purposes.

When the ASCI White supercomputer comesonline at Livermore this summer, it will carryforward the achievements of its predecessor, theBlue Pacific machine, toward the ultimate goalof DOE’s Accelerated Strategic ComputingInitiative—full-scale simulations of nuclearbehavior in support of stockpile stewardship.The article beginning on p. 4 reports both theaccomplishments of Blue Pacific in terascalesimulations and the promise of ASCI White inhelping to fulfill stockpile stewardship’s mission.On the cover is the prize-winning Blue Pacificsimulation of what happens when a shock wavepasses through the interface of two fluids ofdiffering densities. It took 5,832 processors tomake this calculation, the largest, most detailedof its kind to date.

• •

Prepared by LLNL under contractNo. W-7405-Eng-48

Departments

Features

Research Highlights

June 2000

LawrenceLivermoreNationalLaboratory

2 The Laboratory in the News

30 Patents and Awards

Abstracts

23 A Predictable Structure for Aerogels

26 Tibet—Where Continents Collide

ContentsSCIENTIFIC EDITOR

Jean H. de Pruneda

MANAGING EDITOR

Sam Hunter

PUBLICATION EDITOR

Dean Wheatcraft

WRITERS

Arnie Heller, Ann Parker, Katie Walter, and Dean Wheatcraft

ART DIRECTOR AND DESIGNER

Kitty Tinsley

INTERNET DESIGNER

Kitty Tinsley

COMPOSITOR

Louisa Cardoza

PROOFREADER

Carolin Middleton

S&TR, a Director’s Officepublication, is produced by theTechnical Information Departmentunder the direction of the Office ofPolicy, Planning, and Special Studies.

S&TR is available on the Web at www.llnl.gov/str/.

Printed in the United States of America

Available fromNational Technical Information ServiceU.S. Department of Commerce5285 Port Royal RoadSpringfield, Virginia 22161

UCRL-52000-00-6Distribution Category UC-0June 2000

S&TR Staff

3 Accelerating on the ASCI ChallengeCommentary by Michael Anastasio and David Cooper

4 New Day Dawns in SupercomputingWhen the ASCI White supercomputer comes online this summer, DOE’s Stockpile Stewardship Program will make another significant advance toward helping to ensure the safety, reliability, and performance of the nation’s nuclear weapons.

15 Uncovering the Secrets of ActinidesResearchers are obtaining fundamental information about the actinides, a group of elements with a key role in nuclear weapons and fuels.

2 The Laboratory in the News

Lawrence Livermore National Laboratory

Lab signs cooperative contracts with RussiansThe Laboratory signed two contracts recently in Moscow

that will assist Russian weapons experts from the closed cityof Snezhinsk in their transition to civilian employment. (Aclosed city is a formerly secret city where nuclear weaponsresearch was conducted.) The projects include developing oilproduction technology and improving Russia’s fiber-opticcables for the commercial market. Both contracts were signedby representatives of the Laboratory and SPEKTR, a StateUnitary Enterprise.

When oil wells are drilled, they are lined with metalcasings that support the surrounding geology and prevent gas,oil, and water from mixing in the well. SPEKTR, whichalready provides explosive charges for perforating thecasings to allow oil to flow effectively at selected depths willuse the approximately $220,000 in U.S. support over the nextyear to develop perforation technologies that apply to diversegeologic conditions and casings.

Under the second agreement, SPEKTR will raise to worldstandards the quality of its multimode optical fiber,demonstrate production capability to satisfy commercialdemands, and develop relationships with cable suppliers tocommercialize its product.

The two new contracts are part of U.S.–Russian strategicplans for the city of Snezhinsk under the auspices of theNuclear Cities Initiative (NCI), a Department of Energyeffort to help the Russian government provide civilianemployment opportunities to weapons scientists in closedRussian nuclear cities. The goal of the NCI is to enableRussian scientists to remain in their homeland and work onsustainable civilian and commercial projects as facilities inRussia’s weapons complex are downsized or closed.

The All-Russian Research Institute of Technical Physicsalso agreed in principle to form an open computer center atSnezhinsk. Lawrence Livermore and the institute will worktoward a contract to begin a commercial software andscientific computations effort. Skilled Russian softwareengineers working at the center will be able to relieve someof the worldwide shortage of programming talent. High-speed Internet lines will connect the center with customersinside and outside Russia, just as they do at other commercialsoftware development centers around the world.

The Laboratory has also signed an agreement to assist theAvangard plant in the closed city of Sarov in converting fromnuclear weapons manufacturing to the production of kidney

dialysis equipment. A non-Russian firm, which has askednot to be identified for proprietary reasons, will purchaseand market three components manufactured by Avangardfor use in machines distributed worldwide. The hope is thatas a result of this U.S.–Russian agreement, Russian-madeparts and eventually systems will make kidney dialysis moreavailable to Russians.Contact: Bill Dunlop (925) 422-9390 ([email protected]).

Breakthrough in laser glass manufactureA major laser glass milestone has been achieved for the

Laboratory’s National Ignition Facility (NIF) thanks toextensive research and development spearheaded by theLaboratory and two leading high-technology glass vendors.

Schott Glass Technologies, based in Duryea, Pennsylvania,has successfully demonstrated a process to ensure continuousproduction of economical, high-optical-quality, neodymium-doped, phosphate laser glass needed for NIF. A second vendor,Hoya Corp. in Fremont, California, began similar glass-melting operations in April.

Schott has produced more than 20 of the glass slabs neededfor NIF’s demanding optical specifications—at a rate 20 timesfaster than is possible using existing one-slab-at-a-time batch-melting technology.

More than 3,500 laser glass slabs will be needed for NIF.Each slab is about 80 centimeters long, 45 centimeters wide,and 4 centimeters thick and weighs about 37 kilograms.

The costs for developing the continuous melting processhave been shared equally by Livermore and the FrenchCommissariat à L’Energie Atomique (CEA). CEA plans topurchase a similar quantity of slabs for its Laser Megajouleto be constructed later in this decade.

In 1999, Schott and Hoya demonstrated the feasibility ofcontinuous-melt production, but certain glass specificationswere not achieved at that time. In particular, the glasscontained trace quantities of contamination from smallamounts of moisture in the surrounding air and in the initialglass raw materials. And attempts to remove the moisture-derived contamination degraded other glass properties.

Recently, however, Livermore, Schott, and Hoya havecarried out cooperative research aimed at reducing moisturecontamination. Schott first demonstrated the success of thisresearch and the improved technology, which both vendors willuse to manufacture the laser glass for NIF and Laser Megajoule.Contact: Ed Moses (925) 423-9624 ([email protected]).

S&TR June 2000

S fast as computer technology is advancing, the Departmentof Energy’s Accelerated Strategic Computing Initiative

(ASCI) must advance the technology even faster.The reason for the push lies in the inexorable aging of the

country’s nuclear stockpile and DOE’s responsibility to keepthis stockpile viable. On September 25, 1995, President Clintondirected DOE to undertake the necessary activities to ensurecontinued stockpile performance in an era of no nucleartesting, no new weapon development, a production complexwith reduced capacity and capability, and an aging stockpile offewer weapons and fewer types of weapons.

The Stockpile Stewardship Program—of which ASCIrepresents one key component—is DOE’s response to thischallenge. It must provide the tools researchers need to developa detailed understanding of the science and technology thatgovern all aspects of nuclear weapons. It must also proceedquickly so that the necessary tools and scientific understandingare in place within about a decade.

In this race against time, three national laboratories—Lawrence Livermore, Los Alamos, and Sandia—have teamedup with the supercomputing industry to accelerate thedevelopment of high-performance supercomputers. Just asfighter planes regularly break the speed-of-sound barrier,ASCI supercomputers are breaking speed barriers of adifferent sort set by Moore’s law. That is, they are doublingcomputing speeds in terms of teraops (trillions of floating-point operations per second) faster than every 18 months.

The current high-end computer at Livermore is the ASCIBlue Pacific machine built by IBM and delivered in the fall of1998. It was used to perform the first-ever three-dimensionalsimulation of an exploding nuclear weapon primary. Thiscalculation, completed in November 1999, represented the firstsuccessful completion of an ASCI milepost application. Inaddition, this machine has performed a series of first-principlessimulations detailing the molecular interactions of the highlycorrosive compound hydrogen fluoride, which occur in somehigh explosives.

The newest candidate for this innovative lineup is ASCIWhite, scheduled for delivery to Lawrence Livermore this

A

3


Commentary by Michael Anastasio and David Cooper

summer. In terms of pure speed, ASCI White will be at leasttwo and a half times faster than the Blue Pacific machine,which is itself an impressive system. One of the bigtriumphs of ASCI Blue Pacific—which will become routinewith White—is performing detailed three-dimensionalsimulations of complex physical phenomena. Clocking in atmore than 12 teraops, White is the next step in the ASCIplan to produce a 100-teraops system by 2004. One hundredteraops is the entry-level performance needed to performfull-scale simulations of exploding nuclear weapons. Theresults from these incredibly complex three-dimensionalsimulations will be combined with existing nuclear test dataand new, nonnuclear experiments to ensure the safety,reliability, and performance of U.S. nuclear weapons.

The national weapons laboratories and the U.S. high-performance computer industry are not the only entitiesengaged in the ASCI challenge. Through DOE’s AcademicStrategic Alliance Program, the U.S. academic communityalso draws on the power and speed of these supercomputersystems to advance unclassified science-based modelingand simulation technologies applicable to all of ASCI’sresearch areas.

At Livermore, ASCI efforts engage talents across theLaboratory’s organizations, with the Defense and NuclearTechnologies and the Computation directorates providingthe focus and the leadership. This issue’s feature article,beginning on p. 4, highlights the accomplishments of ASCIBlue Pacific and the promise of ASCI White. The progressin computing inherent in these machines has placedresearchers in the DOE community and in academia on theverge of being able to simulate first-principle physicswithout resorting to oversimplified models. It’s an excitingprospect, one that promises breakthroughs not only forstockpile stewardship but also for areas as diverse asbiochemistry, materials science, and astrophysics.

Accelerating on the ASCI Challenge

n Michael Anastasio is Associate Director, Defense and Nuclear Technologies.David Cooper is Associate Director, Computation, and the Laboratory’s Chief Information Officer.

S&TR June 2000

S the beam of a flashlight is to theillumination of a modern sports

stadium, so is the power of an averagepersonal computer to the Department ofEnergy’s Blue Pacific supercomputer.And now, the Blue Pacific, located atLawrence Livermore NationalLaboratory, is about to be eclipsed byanother even higher performancesupercomputer. Dubbed White, themonster machine developed by IBMfor Lawrence Livermore is the latestcomputing system in DOE’sAccelerated Strategic ComputingInitiative (ASCI)—a program whosemission is to produce a system capableof calculating at 100 teraops (trillionfloating-point operations per second)by 2004. “On a logarithmic scale,”says Mark Seager, Livermore’sprincipal investigator for ASCIplatforms, “we’re halfway to that goalwith the 10-teraops ASCI White.”

The Clock Ticks for the StockpileThe push to produce a 100-teraops

machine is tied to the aging of the U.S.nuclear weapons stockpile and thecountry’s commitment to maintaining

4


The ASCI White supercomputer will take DOE

significantly closer to its stockpile stewardship

goal of using full-scale simulations of nuclear

behavior to help ensure the safety, reliability,

and performance of U.S. nuclear weapons.

ANew Day Dawns in SupercomputingNew Day Dawns in Supercomputing

S&TR June 2000

because we’d always replaced oldsystems with new designs before theirretirement dates. And we’d hadunderground nuclear tests to confirmthe viability of the stockpile.”

The answer to that question becameDOE’s Stockpile Stewardship Program.Weapon reliability, safety, surety—allonce verified by underground nucleartests—must now be confirmed usingnonnuclear experiments, historicalunderground nuclear test data, andhigh-fidelity computer simulations.

A key component of the StockpileStewardship Program, ASCI is pushing

5


ASCI White

developing nuclear weapons, theoverriding design question was ‘Wouldit work?’ Back then, we had nucleartesting to prove out the designs. It wasn’treally necessary to be efficient as longas Mother Nature got to vote.” As thenumber of underground tests shrankfrom hundreds to tens per year and thesimulation codes and computers runningthem became more capable, the designprocess for nuclear weapons andsupporting computer codes becamemore efficient. “After the test ban, thequestion became ‘How long?’ Agingnuclear weapons had never been an issue,

and preserving a nuclear deterrent.With the advent of the U.S. nucleartest moratorium in 1992, the rules of thegame changed for the nuclear weaponsstockpile. No longer can the reliability,safety, and performance of these weaponsbe confirmed by underground nucleartests. Originally designed with lifetimesof 20 years or more, these weaponsmust now be kept ready to serve thecountry indefinitely.

Doug Post, associate leader for theLaboratory’s Computational PhysicsDivision, explains, “In the early days,when the Lab was in the business of

One Program—Three Laboratories

The goal of the Accelerated Strategic Computing Initiative(ASCI) is to provide the numerical simulation capability neededto model the safety, reliability, and performance of a completenuclear weapon—from start to finish. The three nationallaboratories working on this initiative are Lawrence Livermore,Sandia, and Los Alamos. Project leaders at each laboratory,guided by the DOE’s Office of the Deputy Administrator forDefense Programs, work with ASCI’s industrial and academicpartners. The overriding challenge is to synchronize the various

technological developments. For example, sufficient platformpower must be delivered in time to run new advanced codes, and networking capabilities must be in place to enable the variousparts of the system to behave as one.

ASCI’s program elements are Applications Development,Platforms, Pathforward, Problem-Solving Environment,Alliances, Visual Interactive Environment for WeaponsSimulation, Verification and Validation, and Distributed andDistance Computing.

S&TR June 2000

computational power far beyond presentcapabilities so weapon scientists andengineers can, with confidence, simulatethe aging of nuclear weapons and predicttheir performance. If high-performancesupercomputing were constrained tocontinue developing at a normal paceas predicted by Moore’s Law (that is,computer speed doubling every 18 months to 2 years), the capabilitynecessary to perform these calculationswould still be a long way from realityby the year 2004. The three nationallaboratories that are part of ASCI—Lawrence Livermore, Los Alamos, and Sandia—have teamed up with thesupercomputing industry to acceleratethe development of high-performancesupercomputing in the marketplace and meet this critical timeline.

The 10-teraops IBM White machineis scheduled to arrive at LawrenceLivermore this summer. With more thantwo-and-a-half times the horsepower ofBlue Pacific, White is an importantrung in the ASCI ladder. “One hundredteraops is the entry level we need to do full-scale simulations of explodingnuclear weapons,” Seager explains.“These simulations must take intoaccount three-dimensional, multiple-physics, high-resolution, coupled

calculations. ASCI White will shedlight on those areas of physics we don’tyet understand, so that when we havethe 100-teraops machine in 2004, we’llhave a better handle on the issues.”

Beating Moore’s LawTo reach its 2004 goal, ASCI

is building the world’s fastestsupercomputers using commerciallymanufactured parts. “Because we needto move faster than Moore’s law allows,”explains Seager, “we’re building oursupercomputers by taking thousandsof processors—basically just like thePC on your desk—and linking themtogether. For ASCI White, we are tyingtogether 8,192 processors to get acorresponding increase in capabilityand speed.”

The processors in White are based on IBM’s latest chip—the Power 3-II.Sixteen processors are grouped into asymmetric multiprocessor node; 512 nodes form the system. Codedevelopers take advantage of the factthat proximity plays an important role inthe speed of communication. Thus, theprocessors within a node can passinformation between themselves quickly,while information between nodes movesa bit slower. Housing and bringing

6


ASCI White

power to this enormous parallel machineare a project in and of themselves. Forthe machine to replace White, LawrenceLivermore is constructing the TerascaleSimulation Facility. (See the box on p. 7.)

As Seager points out, “When we speakof accelerating the technology, we’retalking about more than just the computer.Everything around the computer platformmust also develop at an increased pace—applications, infrastructure, networks,archives, visualization tools. In fact, for every dollar spent on computerhardware, the initiative spends two onsoftware.” (See the box on pp. 10–11.)

Building on Blue’s TriumphsIt’s not easy to develop applications

software for these parallel machines,notes Post. Many of the algorithms usedin past weapons-related codes were notdesigned with parallel computing inmind. Plus, they were written to examinephysics in two dimensions, not three,and had to run on machines much lesscapable than today’s workstations.Consequently, simplified assumptionsabout the physics processes and thegeometries involved are part of thosecodes. To produce the high-fidelitysimulations required for stockpilestewardship, ASCI codes must remove

1996 1998 2000 2002

Time

Weapon aging

Designer test experience

ASCIcomputationalrequirement

~20 to 40 years

2004 2006 2008 2010

The year 2004 is a critical one for theStockpile Stewardship Program. By then,almost all of the country’s nuclear weaponsystems will be at least 20 years old—theirtypical design lifetime—and most of theweapon scientists and engineers whoworked on weapons during the nuclear-testera will have reached retirement age.

these assumptions and replace them withmuch more accurate methods based, inmany cases, on a first-principles approach.

As part of ASCI, the Laboratory’scode developers set about writing codesto take advantage of the architecture ofthis new generation of supercomputers.Simulation codes for ASCI includenuclear performance codes that canpredict the details of an explodingnuclear weapon, materials modelingcodes that use molecular dynamics to

study the long-term degradation ofmaterials under the influence of low-level radiation, and multiphysics codesthat simulate what happens within aninertial confinement fusion target duringimplosion. Many of these codes haverun on the ASCI Blue Pacific machinewith impressive results.

Modeling a PrimaryIn one notable ASCI Blue Pacific

triumph, Lawrence Livermore, with

strong support from IBM, completedthe first major ASCI scientific milestonein November 1999—the first-ever three-dimensional simulation of an explosionof a nuclear weapon primary. (Nuclearweapons have two main components:the primary or trigger, and the secondary,which produces most of the energy.)Demonstrating the ability tocomputationally simulate, visualize, and analyze what happens to each of a nuclear weapon’s components is a

7


ASCI WhiteS&TR June 2000

The Terascale Simulation Facility

The time scales and demands of ASCI require a facility ofextraordinary scope: the Terascale Simulation Facility (TSF).The $89-million facility will house a computer complex and dataassessment and networking capability. It will require a significantincrease in electrical power, mechanical support, physical space,and networking infrastructure. The TSF will provide an acre of

raised computer floor and a power plant equivalent to thatservicing a city of 15,000 homes.

ASCI White, with a delivery date of mid-2000, will residenext door to the proposed TSF in Building 451, where power andcooling systems have been tripled to meet the requirements of the10-teraops supercomputer.

critical step in simulating an entirenuclear weapon’s explosion in threedimensions.

The simulation required about300,000 megabytes of random-accessmemory (RAM). For comparison, aconventional desktop computertypically comes with only about 100 megabytes of RAM.The calculation—

Just looking at this much informationpresents a formidable challenge.

Tom Adams, an associate divisionleader in Defense and NuclearTechnologies Directorate, explains,“We had so much data that we neededto use the ASCI machine to analyze theresults as well.” ASCI’s visualizationsoftware was also put to the test. Resultswere displayed as movies, in whichscientists could look inside the primaryto see what was happening at differentpoints in time, and as graphic renditionsof temperatures, pressures, and so on.

“This burn code ‘milepost’ wasoriginally set in 1996 as a target to hitbefore the year 2000,” continuesAdams. “It was a daunting objectivethat some thought we’d never meet. In accomplishing it, we’ve shown thatthe ASCI program is on track, that thehardware and software systems areworking, and that we can bring in results.The run also tested the teamwork ofthe people from the physics code teams,the Computation Directorate, and IBM.They all worked hard to solve theproblems and make this happen. Wesee this exercise on Blue Pacific as a model for what we plan to do with White.”

Adams expects that with White, thesimulations will get more detailed. “We’llbe able to model a larger part of the fullweapon system, input more details onthe weapon’s configuration, and usemore complex coupled physical models.All of this will improve the fidelity ofour simulations.”

The next ASCI scientific milestoneis to model the secondary, somethingthat will require White’s increased powerand speed. “In addition, we’re movingfrom this, our first complete calculation,to something that nuclear weaponscientists and engineers can use foranalyzing the stockpile,” says Adams.“We’re working with these users toapply the codes to current stockpilestewardship issues.”

8


ASCI White S&TR June 2000

ASCI Blue Pacific wasused to create thelargest-ever three-dimensional directsimulation of neutrontransport. The colorsrepresent flux values—with red highest andblue lowest—of fusionneutrons coming out ofthe Nova laser targetchamber.

which would have taken 30 years on adesktop—ran for more than 20 days onthe Blue Pacific using 1,024 processors.The complex computer model, referred toas the burn code, employed tens ofmillions of zones—hundreds of timesmore than a comparable two-dimensionalsimulation using traditional codes. The

calculations produced 50,000 datafiles containing 6 terabytes

of numericalinformation.

A simulation of the the fluid motion as a function of increasing temperature, pressure, anddensity (a Richtmyer–Meshkov instability) in an imploding inertial confinement fusion capsulecalculated with an arbitrary Lagrange–Eulerian hydrodynamics code on ASCI Blue Pacific.

Dissecting ICFSeveral of the codes developed for

ASCI examine various physics processesrelated to inertial confinement fusion(ICF). For instance, one three-dimensional code examines what occursinside an ICF pellet when x rays hit thepellet’s surface. The code models thefluid motion within the capsule astemperature, pressure, and densityincrease. It also models the transport ofx rays emitted by the high-temperaturematerial and the energy released from thefusion process. Another effort focuses onmethods for modeling radiation transport.Seager notes, “Radiation transport iscentral to many physics applications,including nuclear weapons, inertialconfinement fusion, plasma processing,and combustion. Many productioncodes in ASCI spend about 80 percentof their time claculating radiationtransport. Doing these calculationsmore efficiently is a big time-saver.”

Another parallel code, which ran onthe ASCI Blue Pacific in 1999, simulatedthe flux of fusion neutrons that comes outof the Nova laser target chamber. Thishigh-fidelity simulation had to take into

account complex three-dimensionalgeometries, a wide range of distances(from the 6-meter-diameter test chamberto a target pellet less than 1 millimeterin diameter), and a number of differentmaterials (air, aluminum, gold) withdensities varying by a factor of 108. This calculation employed more than160 million zones with over 15 billionunknowns and took 27 hours to solve on3,840 processors.

9



(a) A volume rendering ofentropy created at theconclusion of the largestcalculation ever run of aRichtmyer–Meshkov instability.(b) and (c) The high resolutionof the three-dimensional ASCIsimulations revealed fine-scalephysics of the turbulence neverseen before. This 8-billion-zonesimulation was completed in justover a week; over 2 terabytes ofgraphics data were produced inmore than 300,000 files. (Thisand other visuals and moviescan be viewed at www.llnl.gov/CASC/asciturb/simulations.html.)

(a)

(b) (c)

10



Tomorrow’sSupercomputer Today 1Whether modeling supernovas or

exploding nuclear weapons, thesimulation starts with the softwaretaking an object and dividing it intobillions of smaller cells or zones.Sometimes, these gridded zonesare cubical; sometimes they takeother shapes such as hexagons.The behavior of each zone isgoverned by a set of physicsequations and numbers that definethe initial condition of the simulation.The initial conditions stream out ofrandom access memory (RAM) and proceed to a switch.

2The switch is the heart of thesupercomputer. The switch moves dataamong ASCI White’s 8,192 processorsand 10,752 external disk drives. Thedata will move from each node (a groupof 16 processors) at 800 megabytes persecond—more than 5 times faster thanBlue Pacific’s switch.

As with Blue Pacific, ASCIWhite will perform incrediblycomplex calculations by dividingup programs so that they will runsimultaneously on thousands ofprocessors. The supercomputercombines the resulting data of anevent, creating a kind of three-dimensional movie from the basiclaws of physics.

From Shock Waves to BuckyballsIn other first-time-ever Blue Pacific

calculations, researchers explored suchdiverse areas as turbulence, ab initiomolecular dynamics, and quantumchemistry.

What happens when a shock wavepasses through an interface of twofluids of differing density? A detailedsimulation on 5,832 processors aimedat answering this question netted ateam from Lawrence Livermore, theUniversity of Minnesota, and IBM theprestigious Gordon Bell Award lastNovember. The simulation, the largestcalculation of its type, achieved agreater level of detail than any previousturbulence simulations. The effort hasapplications in a variety of disciplines,including supersonic propulsion,combustion, and supernova evolution.

Researchers at the Laboratory used3,840 processors to simulate from firstprinciples the molecular interactions of hydrogen fluoride, an extremely toxicand corrosive byproduct of insensitivehigh-explosive detonations. Littleexperimental data are available onhydrogen fluoride, particularly at hightemperatures and pressures. Quantummolecular dynamics simulated theinteraction of hydrogen fluoride andwater at the microscopic level—the onlyinput being the identities of the atomsand the laws of quantum mechanics. ThisASCI simulation involved 600 atomswith 1,920 electrons. The simulationsprovided crucial insight into the propertiesof hydrogen fluoride–water mixtures athigh pressures and temperatures, addingto the understanding of how insensitivehigh explosives perform. (See S&TRJuly/August 1999, pp. 4–11, for moreinformation about quantum moleculardynamics and this simulation.)

A research collaboration fromLawrence Livermore, the University ofCalifornia at Berkeley, and SandiaNational Laboratory at Livermore usedASCI Blue Pacific to perform thelargest first-principles quantum chemistrycalculations ever done. One of the initial

http://www.llnl.gov/str/Haan.html

S&TR June 2000 11


ASCI White

3In ASCI White, each 375-megahertz Power 3-IIprocessor simultaneously executes four floating-pointcalculations and contains 8 megabytes of cache RAM.Sixteen processors make up a single node, comparedwith the IBM Blue Pacific, which had four processors pernode. Each White node also contains about 8 gigabytesof local memory and two internal 18-gigabyte hard drives.

4As zone data enter the nodes,they flow into local memory andare distributed among theprocessors. The processorsperform mathematicalcalculations and produce resultsfor each zone. The zone resultsare sent back through the switchand stored in the “disk farm.”

5The massive amountsof data—hundreds oftrillions of zones—arestored in 195 terabytesof external disk storage(for comparison, all theprinted material in theLibrary of Congress isabout 10 terabytesworth of information). 6By means of ASCI’s special

visualization software, theresults are displayed in three-dimensional movies, cutawayviews, and graphs showing thedistribution of density, pressure,temperature, and otherquantities needed tounderstand the calculation andits implications.

applications was to determine the three-dimensional structure and electronic stateof the carbon-36 “buckyball,” one of thesmallest, most stable members of thebuckminsterfullerene family ofcompounds. Possible applications forthese unusual compounds include high-temperature superconductors and precisedelivery of medicines to cancer cells.Previous quantum chemical calculationsnarrowed down the possible structure ofcarbon-36 to one of two possibilities,shown below right. Experimental datafavored the structure in (a), buttheoretical results slightly favored theone in (b). The two structures havedifferent chemical properties, sodetermining which is more stable wascritical to understanding this compound.A high-fidelity ASCI calculation providedthe definitive—and unexpected—answer.As it turned out, the structure in (a), theone favored by the experimentalists, wasthe most stable. “This exercise was areminder of how low-fidelity simulations

with their simplifications andinterpolations can lead to catastrophicresults,” notes Seager. “In a weaponscalculation, for instance, you don’t wantto get the wrong answer to the question‘Will it work?’”

Universities Logging OnSignificant scientific work on ASCI

Blue Pacific is proceeding not only atthe national laboratories, but also atassociated universities through DOE’sAcademic Strategic Alliance Program.This program aims to engage the U.S.academic community in advancingscience-based modeling and simulationtechnologies. Although the specificcomputing problems universities aretackling do not directly involve nuclearweapons research, the methodologiesand tools being developed can be appliedto all of ASCI’s areas of research.

The program funds five major centersof excellence. Each uses multidisciplinaryteams working over the long term to

provide large-scale, unclassifiedsimulations that represent ASCI-classproblems. The centers collectively haveaccess to up to 10 percent of the ASCIcomputing resources at the three nationallaboratories. The projects are part of a10-year program, in which projectscome up for renewal after five years.

“The problems being studied inthese projects are comparable in theircomplexity to those involving nuclearweapons,” explains Dick Watson,Lawrence Livermore’s manager for theprogram. “We expect that through thisprogram, the laboratories and universitieswill see revolutionary advances in boththe physical and engineering sciences andthe mathematical and computer sciences.”

12



Two possible structures for the carbon-36fullerene. It took a high-fidelity run on ASCIBlue Pacific to determine that the correctstructure is (a).

This quantum-levelsimulation of hydrogenfluoride–water mixturesat high temperaturesand pressures took 15 days on the ASCIcomputer. Trillions ofoperations per secondwere performed tocalculate 1 picosecond’sworth of atomicinteractions.

(a)

(b)

The Center for Simulating DynamicResponse of Materials, based at theCalifornia Institute of Technology, isdeveloping simulation codes for its virtualshock physics test facility. Simulationshave applications to materials design,oil exploration, earthquake prediction,and environmental analysis.

Stanford University’s Center forIntegrated Turbulence Simulationsfocuses on jet-engine simulations.Researchers at this center are developingmassively parallel codes for high-fidelityflow and combustion simulation.

The Center for AstrophysicalThermonuclear Flashes at the Universityof Chicago is studying the physics ofsupernovas, including the physics ofignition, detonation, and turbulent mixingof complex fluids and materials. Oncecompleted, the integrated code willprovide the highest resolution calculationever done showing how these stellaroutbursts begin.

The University of Illinois Urbana–Champaign is home to the Center forSimulation of Advanced Rockets. Thiscenter plans to provide a detailed, whole-system simulation of solid propellantrockets. Earlier this year, center staffcompleted a simplified version of a three-dimensional integrated rocket simulationcode. The next-generation code willcharacterize various burn scenarios,and the fully integrated code willaddress potential component failures.Research will benefit technologies suchas gas generators used for automobileair bags and fire suppression.

Finally, the University of Utah isfocusing on the physics of fire at theCenter for the Simulation of AccidentalFires and Explosions. These problemsdraw on fundamental gas- and condensed-phase chemistry, structural mechanics,turbulent flows, convective and radiativeheat transfer, and mass transfer.

At Lawrence Livermore, all of thealliance work is being conducted on theunclassified sector of ASCI Blue Pacific.

Building a Model for the Future“One of the key challenges to fielding

the fastest supercomputers in the world isthat this scale of computing requires anew operational model in order tosucceed,” says Seager. “In the past, themodel has been a large, but fundamentally

traditional, scientific computing center.ASCI needs to be run as if it were anexperimental research facility. Thescientific applications being developedtoday promise a level of physical andnumerical accuracy that is more like thatof a scientific experiment than a

13



A simulation of the flow through a compressor in a jet engine, showing entropy contours. Red ishigh entropy; blue is low.

A snapshot of a densityfield simulation for an x-ray burst on the surfaceof a neutron star. Theyellow curve is thedetonation front, racingacross the stellarsurface. The blue curveshows how the initialsurface of the accretedatmosphere deforms.

traditional numerical simulation. Theeffort required to run a full-scale, three-dimensional scientific application is likerunning a big experimental weapon test.In a way, this operational change parallelsthe changing role of large-scaleexperiments in the physics world.”

For a long time, the two acceptedbranches of physics have been theory andexperiment. Yet, there are experimentsthat, for various reasons, are unrealistic—either the conditions can’t be created inthe laboratory, or it’s far too dangerousor expensive to do so. And, in manycases, the theory has been too complex to analyze, even with many simplifiedassumptions. “All that has changed withthe coming of ASCI’s supercomputers,”says Seager. “Computer simulation isnow poised to become the new branch of science, on the same level asexperimentation and theory.”

—Ann Parker

Key Words: Academic Strategic AllianceProgram, Accelerated Strategic ComputingInitiative (ASCI), ASCI White, Blue Pacific,Stockpile Stewardship Program, TerascaleSimulation Facility (TSF).

For more information contact Mark Seager (925) 423-3141([email protected]).

Also see the following Web sites:• ASCI at Lawrence Livermore,www.llnl.gov/asci/• ASCI at Los Alamos, www.lanl.gov/asci/• ASCI at Sandia, www.sandia.gov/ASCI/• ASCI’s Academic Strategic AlliancesProgram, www.llnl.gov/asci-alliances/

14



MARK SEAGER is the principal investigator for the AcceleratedStrategic Computing Initiative’s terascale system at LawrenceLivermore National Laboratory. His interests are currently in large,scalable system architecture, performance, and integration. Seagerreceived his B.S. in mathematics and astrophysics from theUniversity of New Mexico at Albuquerque in 1979 and a Ph.D. innumerical analysis from the University of Texas at Austin in 1984.

About the Scientist

A three-dimensional simulation of the gas temperature at about 0.04 seconds after rocket-propellantignition. Orange represents the hottest region. (Based on simulations of gas flow and pressuresinduced by the shape of the propellant inside the ignition segment of a Space Shuttle booster.)

http://www.llnl.gov/asci/

http://www.lanl.gov/asci/

http://www.sandia.gov/ASCI/

http://www.llnl.gov/asci-alliances/

15


S&TR June 2000

A family of radioactive

elements, the actinides, is

key to safe stewardship of

nuclear weapons.

Uncovering the Secrets of Actinides

NDERSTANDING the periodic table, withits assemblage of columns and rows of

elements, has been a perennial challenge forchemistry students. (See the box on p. 17.)Understanding at the atomic level a remarkablerow of elements has been a particular researchchallenge for Lawrence Livermore scientists overthe years. That row is called the actinides, acollection of 14 radioactive elements named afterthe element actinium.

“There’s a tremendous amount we don’t knowabout the actinides,” says Lawrence Livermorechemist Lou Terminello, who leads the MaterialsScience and Technology Division of the Chemistryand Materials Science Directorate. To learn moreabout these elements, he says, the Department ofEnergy funds about $100 million per year forresearch at Lawrence Livermore. The research isconducted by teams of chemists, physicists,engineers, metallurgists, and environmentalscientists on a diverse set of national security

and environmental issues.Terminello says that

a more fundamental understanding of actinides is needed to better assess the

nation’s nuclear stockpile, help stem the clandestine proliferation of nuclear

weapons, and better understand the implicationsof nuclear fuels’ (such as enriched uranium) useand storage. Environmental contamination byactinides is also a major concern at several majorDOE facilities. In addition, actinides such as

U

uranium, neptunium, plutonium, andamericium are the major contributors tothe long-term radioactivity of nuclearwaste currently targeted for the proposedYucca Mountain repository in Nevada.

Stockpile stewardship, DOE’sprogram for certifying the long-termsafety and performance of the enduringstockpile without underground nucleartesting, has heightened the importanceof assessing and predicting the long-term behavior of actinides. A majorfocus is on obtaining a better scientificunderstanding of the isotopes uranium-235 and, especially, plutonium-239.

Plutonium is the most complex andperplexing element in the periodic table.The element’s complexity stems in partfrom its mercurial nature. Depending on temperature, it assumes one of sixdifferent forms or phases, each with adifferent density and volume. Becauseof plutonium’s enigmatic behavior and the need for stringent safety andenvironmental procedures when

handling the toxic material, much ofthe extensive characterization workdone on other metals has not beenperformed on plutonium.

Surrogates InadequateMaterials scientist Mike Fluss points

out that because of plutonium’sunpredictability, experimenters prefernot to use surrogate materials. “It’s aschallenging a material as you canimagine,” he says. Even the process ofmeasuring its electrical resistance hasproven surprisingly complex because ofits unexpected and not fully understooddependency upon temperature.

“We’re rebuilding plutonium metalsscience at Lawrence Livermore,” saysmetallurgist Adam Schwartz. He pointsto a growing number of experimentsmeasuring the structural, electrical, andchemical properties of plutonium and itsalloys and determining how they changeover time as a result of the cumulativeeffects of radioactive decay and

16


S&TR June 2000Actinides

IA

IIA

4

Be

IIIB IVB VB VIB VIIB

VIIIB

IB IIB

IIIA IVA VA VIA VIIA

0

3

Li

1

H

11

Na

19

K

37

Rb

55

Cs

87

Fr

88

Ra

39

Y

21

Sc

104

(Rf)(261)

72

Hf 178.49

40

Zr

22

Ti

105

Db

107

Bh

108

Hs

109

Mt

110

111

112

114

116

118

73

Ta

41

Nb

23

V

106

Sg

74

W

42

Mo

24

Cr

75

Re

43

Tc

25

Mn

76

Os

44

Ru

26

Fe

77

Ir

45

Rh

27

Co

78

Pt

46

Pd

28

Ni

79

Au

47

Ag

29

Cu

80

Hg

48

Cd

30

Zn

81

Tl

49

In

31

Ga

82

Pb

50

Sn

32

Ge

83

Bi

51

Sb

33

As

84

Po

52

Te

34

Se

85

At

53

I

35

Br

86

Rn

54

Xe

36

Kr

13

Al

14

Si15

P

16

S17

Cl

18

Ar

5

B6

C7

N8

O9

F

10

Ne

2

He

56

Ba57

La

38

Sr

92

U(238)

93

Np (237)

94

Pu(244)

95

Am(243)

96

Cm(247)

97

Bk(247)

98

Cf(251)

99

Es(252)

100

Fm(257)

101

Md(258)

102

No(259)

103

Lr(260)

91

Pa(231)

90

Th(232)

20

Ca

12

Mg

Actinides

Thorium

Protactinium

Uranium

Neptunium

Plutonium

Americium

Curium

Berkelium

Californium

Einsteinium

Fermium

Mendelevium

Nobelium

Lawrencium

89

Ac(227)

The elements from actinium (element 89) to lawrencium (element 103) form a distinctgroup—the actinides—within the periodic table.

consequential damage. Thesemeasurements will enable scientists tobetter model and predict the material’slong-term behavior in the nation’s agingnuclear stockpile.

Schwartz also cites the recentacquisition of advanced instrumentssuch as a transmission electronmicroscope capable of nearly perfectresolution at the atomic scale.Additionally, Livermore experts aretaking advantage of one-of-a-kindfacilities at Lawrence Berkeley andArgonne national laboratories, theStanford Linear Accelerator Center, and other DOE sites to more completelycharacterize the electronic and atomicstructure of plutonium alloys andcompounds.

One line of research is studying theevolution of damage to plutoniummetal’s crystalline structure on scales assmall as a billionth of a meter. This so-called microstructure is alwayschanging because when plutonium-239decays, it emits a 4-megaelectronvoltalpha particle (a helium nucleusconsisting of two protons and twoneutrons) and an 85-kiloelectronvoltrecoiling atom of uranium-235. Theresulting buildup of gaseous heliumatoms and displaced plutonium atomsfrom the recoiling uranium couldproduce unacceptable changes in the

plutonium metal. Fluss notes that after10 years, every plutonium atom has beendisplaced at least once from its latticesite, but most atoms eventually returnthere. The plutonium decay itself isslow; in about 24,000 years, only halfthe plutonium-239 has changed touranium-235.

The concern, says Fluss, is that atomsof helium and the actinides americiumand uranium, also present in the weaponenvironment, might slowly change thechemistry of the plutonium metal. At thesame time, the accumulation of small-scale radiation damage to plutoniumalloys over several decades could affecta weapon’s safety or its performance.Like other solids, plutonium metal ismade of many crystals (or grains) withdifferent orientations. If vacancies ordefects coalesce, they may causechanges in properties, with possibleunwanted effects to a warhead. Bybetter understanding the nature of thechanges, scientists can refine theirpredictive codes.

100-Year Predictions Needed“We need to know how plutonium in

our stockpile will react over 100 years,”says Fluss. “We’re asking harderquestions today because nuclear weaponsmust last a lot longer than theirdesigners ever intended.” The answer,he says, lies in obtaining fundamentalunderstanding at the atomic level.

Schwartz and colleague Mark Wallare using the transmission electronmicroscope to document thedifferences between plutonium fromold, disassembled nuclear warheadsand newly cast plutonium. By usingelectrons instead of light waves, thetransmission electron microscope canimage features at near-atomic resolution.They start with plutonium samplesmeasuring less than 3 millimeters indiameter and 120 micrometers thick.The center of each sample is thinned tocreate a region only 100 nanometers

17


S&TR June 2000 Actinides

Metallurgists are probing the microstructure of actinides with a transmission electron microscopecapable of near-atomic resolution.

Actinides Can Mean Nuclear Chemistry

The group of elements known as theactinides are the elements from actinium(element 89) to lawrencium (element 103).All members of the series can resembleactinium in their chemical and electronicproperties, and so they form a separategroup within the periodic table. (Anelement’s atomic number is the sum ofthe protons and neutrons in the nuclei of its atoms.)

All actinides are metals and all areradioactive. As a result, they dominate thestudy of nuclear chemistry. The elementsemit energy in the form of alpha particles,beta particles, or gamma rays. By emittingthese particles, the atoms lose protonsand therefore become another elementwith a lower atomic number. If theimmediate product of radioactive decayis radioactive, it also decays to formanother element. This process continuesuntil a stable element is formed.

Actinides undergo radioactive decayat different rates; that is, they have differenthalf-lives. Elements with higher atomicnumbers have short half-lives and rapidradioactive decay. Some actinides withlower atomic numbers, however, havehalf-lives ranging between thousands tomillions of years.

The two actinides of most interest toLivermore scientists are uranium andplutonium. Uranium, a silver andlustrous metal, has four main isotopes.Because uranium-235 is fissionable, it isused to fuel nuclear power plants and asa component in nuclear weapons.

Plutonium is a silver-gray metal thathas 16 isotopes. The isotope of chiefinterest is plutonium-239, which, likeuranium-235, is fissionable. Most nuclearweapons are based on plutonium-239,while plutonium-238 is used as a powersource in long-mission space probes.

(100 billionths of a meter) thick for theelectron beam to pass through. Theresulting electron micrographs revealthe nature and extent of defects inunprecedented detail.

Researchers are also doing a varietyof accelerated aging studies in whichplutonium samples are exposed to higherthan normal levels of radiation so thatthe aging process is significantlyaccelerated. Such experiments providean important basis for validatingcomputer models. Other aspects of thephysics of radiation effects are beingstudied by ion irradiation using variouslight and heavy ions to investigate thepredictions of the models.

Lawrence Livermore scientists arealso benefiting from fundamental work onplutonium performed by their colleaguesin Russia. One study, done over the past25 years and announced last year,claims to have produced plutonium’scorrect phase diagram, the roadmapbetween its six different phases orstructural forms. The Russian study,says Schwartz, clarifies certain detailsabout how delta-phase plutoniumtransforms to a less desired alpha state.

Fluss, Schwartz, and others areplanning research that will tap theLaboratory’s resources to review theRussian work and test its conclusions.

The likely outcome, says Schwartz, is arefinement of current computer codesto more realistically simulate the natureof plutonium.

New Look at Old DataOther Lawrence Livermore

researchers are taking a differentapproach to strengthening the ability of scientists to predict the likelyperformance and safety of agingweapons. The scientists are looking at results from years of undergroundnuclear detonations at DOE’s NevadaTest Site.

According to nuclear chemist KenMoody, new measuring techniques and instruments, along with improvedunderstanding of actinide chemistry,warrant revisiting test data that aredecades old. Moody is one of a dwindlingnumber of nuclear chemists who did theoriginal chemical separation of actinidesfrom underground tests before theyceased in 1992. He notes that storedactinide samples and even debris fromtests could be a treasure-trove of data,despite their reduced radioactivity dueto age. The reanalysis could givestockpile stewards a clearer idea of howthe nuclear devices performed when theywere detonated and how those samedesigns would perform today.

Livermore researchers are applyingtheir actinide know-how and a suite of sensitive instruments to nuclearforensics work. Chemists like Moodyare working with Lawrence Livermore’sForensic Science Center to helpAmerica’s intelligence agencies stemthe proliferation of nuclear materials,especially those from the former SovietUnion. Experts have raised concern aboutthe security of large amounts of weapons-grade nuclear materials in Russia andneighboring states that inherited thematerials as a result of the breakup ofthe Soviet Union. In particular, thedismantlement of thousands of old Sovietnuclear weapons has resulted in largequantities of surplus nuclear materials.

Some actinides, such as uranium-235(used in nuclear fuel rods) andplutonium-239, have shown up in smallquantities in unauthorized hands and onblack markets in Western Europe. Theconcern, of course, is that such materialsmight make their way to a terroristgroup or a nation that supports terroristactivities.

Lawrence Livermore’s actinideforensics capabilities are formidable,Moody says. Radiochemical methodscan reveal, for example, when a sampleof plutonium was manufactured andeven the chemical techniques used in its

18



Images of plutonium metaltaken with the transmissionelectron microscope revealchanges to the plutonium’smicrostructure. At the 0.4-micrometer scale,diagonal bands in (a) aretypical of accumulations of a deformed microstructure.Dark lines in (b) at the 100-nanometer scale areindividual dislocations.

(a) (b)

100 nanometers0.4 micrometer

19


ActinidesS&TR June 2000

Educating Future Actinide Scientists

The Department of Energy nationallaboratories have long been the stewards ofexpertise in actinides in the United States.However, many actinide experts are retired or in the process of retiring, and they are not being replaced in adequate numbers.

The situation largely results from a sharpdownturn in the number of students graduatingwith specialties in nuclear chemistry. Across the nation, only a few colleges and universitiesstill provide facilities for actinide research, and professors teaching actinide science havemostly retired. Also, fewer undergraduates are expressing an interest in pursuing careers in nuclear chemistry.

According to Livermore actinide experts, U.S.leadership in heavy-element science will fast erode unless thenational laboratories address this issue, which is vital to DOEstockpile stewardship and other missions such as nuclear wastedisposition. “One of the most important challenges facing stockpilestewardship is the successful passing of the torch in actinidescience,” says Lawrence Livermore chemist Lou Terminello.

The University of California’s Glenn T. Seaborg Institute forTransactinium Science is attempting to remedy the labor shortage byattracting and training the next generation of actinide scientists. Theinstitute was established in 1991 with facilities at both LawrenceLivermore and Lawrence Berkeley national laboratories. (A thirdchapter was added in 1997 at Los Alamos National Laboratory.)

The institute is named for the late UC Berkeley professor inrecognition of his enormous contributions to the field, includingthe discovery of 10 elements, among them plutonium. Theinstitute advances fundamental and applied science andtechnology of transactinium elements (actinides and beyond).Workshops, conferences, lectures, and research projects focus onnational security, nuclear energy, environmental protection andremediation, and nuclear waste isolation and disposition.

The institute emphases training at the undergraduate throughpostgraduate levels. In this way, says Terminello, who serves asinstitute director, Lawrence Livermore is making a long-term investment in its future. To that end, theinstitute’s Livermore facility operates a summer school forundergraduates who have shown an interest in nuclear chemistry.“We want to capture the imagination of young people by givingthem hands-on experience in nuclear science. We want them to go to graduate school and return to Livermore, where they willform our next generation of actinide scientists,” Terminello says.

Performing research on actinides for stockpilestewardship often requires training beyond that whichis available from universities. As a result, the institutealso trains chemists who have recently obtained a Ph.D.For example, young scientists are learning thetechniques of x-ray absorption that were refined foractinides by Livermore chemists such as Patrick Allen,deputy director of the institute.

The researchers use the facilities of the StanfordSynchrotron Radiation Laboratory, a part of theStanford Linear Accelerator Center. The laboratorygenerates synchrotron radiation, a name given to x rays produced by electrons circulating in a storagering at nearly the speed of light. The extremely bright x rays excite electrons closest to the nucleus, yieldingdetailed information about the chemical nature,

molecular structure, and electron distribution of actinide-containingmaterials.

Lawrence Livermore researchers use x-ray absorption to probesamples of uranium and plutonium alloys and compare the results tocurrent computer models. The results are useful in addressingstockpile stewardship issues as well as understanding the behavior ofactinides in contaminated soils and potential radioactive wastestorage facilities.

Whatever theexpense of improving

education, it is aninvestment in thefuture we must

make. Excellencecosts. But in the longrun mediocrity costs

much more.—Glenn T. Seaborg

Professor Glenn T. Seaborg poses with college students participatingin the first summer session (1998) at the Glenn T. Seaborg Institutefor Transactinium Science at Lawrence Livermore.

creation. They can also readily show ifa suspect material is a hoax rather thana real threat.

Creating Element 114The accumulated knowledge of

actinides’ nuclear structure has helpedLawrence Livermore scientists createentirely new elements. In 1989, aLivermore team led by nuclear chemistKen Hulet (now retired) began acollaboration with scientists at the JointInstitute for Nuclear Research in Dubna,Russia. Over the past decade, theinternational team discovered isotopesof elements 106, 108, and 110 at theRussian institute.

The researchers’ goal in 1998 was farmore challenging: to create element 114and demonstrate a long-postulated regionof enhanced nuclear stability against

20



A Russian artist depicts the modern nuclear theory of the heaviest elements. At the upper far right is the island of stability, which was demonstratedby the production of long-lived element 114 in 1998 by a Russian–Livermore team.

For more than a decade, Russian and Lawrence Livermore collaborators have used the facilitiesof the Joint Institute for Nuclear Research in Dubna, Russia.

114

100

90

126 142 146 162 184Neutron number

Pro

ton

num

ber

Shoal ofdeformed

nuclei

UraniumThalium Plutonium

Mountains

Lead

Continent

Island of stabilityof superheavy

spherical nuclei

spontaneous fission. This region,considered by some impossible to reach,was theorized to lie amidst a “sea” ofextremely short-lived, super-heavy nuclei.

The most recent experiment, led byMoody, involved a team of fiveLivermore scientists and 17 Russianresearchers. The team bombarded ionsof the rare isotope calcium-48 onto atarget of plutonium-244 (the heaviestlong-lived plutonium isotope) suppliedby Livermore. It took the team 40 daysof irradiation to create one atom of thenew super-heavy element 114 inDecember 1998. The new elementlasted 30 seconds, some 100,000 timeslonger than if there were no enhancedstability in that area of the periodic table.

Moody believes the discovery ofelement 114 has strengthened interestin heavy-element science. Since thediscovery, Lawrence Berkeley researchershave found two new elements—116and 118. The continuing discoveriesare providing important insights intothe arrangement of electrons in atomsand chemical bonding.

The Livermore–Russia connection isstill going strong. Since the December1998 discovery, the U.S.–Russian teamhas found a different isotope ofelement 114, one with a decay time of one second. The team is currentlyworking on finding another isotope ofelement 116, this time with a curiumisotope target, so they can continuemapping the region of enhanced stability.

Plutonium Moves DifferentlyAn important aspect of Lawrence

Livermore actinide research is studyinghow these elements behave in theenvironment, particularly how theymigrate underground in solution. Theresearch results have challenged somelong-established scientific assumptions.

For example, scientists assumed thatplutonium, because of its low solubilityin water and its strong tendency to sorb(adhere) to clumps of dirt and rocks,does not travel far in groundwater. ALawrence Livermore–Los Alamos team

led by Livermore geochemist AnnieKersting has shown, however, thatplutonium can adhere to colloids, whichare naturally occurring particles of rocksmaller than a micrometer in diameter.In this way, small amounts of plutoniumcan be transported considerabledistances by groundwater.

The team studied the distancesplutonium ions had traveled from thePahute Mesa region of DOE’s NevadaTest Site. The group analyzed some ofthe groundwater pumped from two deepsampling wells dug near the sites wherefour underground nuclear tests had beenconducted. The researchers discoveredthat colloids filtered from the watercontained more than 99 percent of thesmall amount of plutonium found in thewell-water samples. (In contrast,99 percent of the tritium was found insolution, and virtually none was foundin the filtrates.) The team proposed that

small amounts of plutonium had adheredto mineral colloids that were transportedby groundwater away from the testlocation.

The team ascertained which of thefour tests conducted in the area hadproduced the plutonium by measuring theratio of the plutonium-240 isotope to theplutonium-239 isotope. (Every nucleartest can produce a unique ratio of the twoplutonium isotopes.) The isotopic ratiomeasured on the groundwater colloidsmatched that of the 1968 Benhamunderground test, which was conducted1.3 kilometers from one of the wells.

“We were surprised to find that theplutonium in the wells was from theBenham test because 1.3 kilometers is along distance for plutonium to migrate,”says Kersting. She adds, however, thatthe detected plutonium concentrationwas extremely small and did not pose a health risk.

21


S&TR June 2000 Actinides

Groundwaterpumped from wellsdrilled in PahuteMesa at theNevada Test Siteshowed that smallamounts ofplutonium traveledsurprising distancesunderground.

The team’s findings have importantimplications for the proposed YuccaMountain nuclear waste repository inNevada (see S&TR, March 2000,

22



pp. 13–20.) The findings may also beapplicable to other DOE sites such asRocky Flats in Colorado and the HanfordNuclear Reservation in Washington,although their underground geologydiffers from the Nevada Test Site’s.

In light of the team’s research,Kersting says that models that do notallow for transport of plutonium bycolloids may significantly underestimatehow far and fast the element can travel.She also notes that colloids may beimportant in the transport of otheractinides. “We want to know howother actinides such as neptunium,americium, and uranium moveunderground,” she says.

Kersting is collaborating with otherLivermore geochemists to determineexperimentally how actinides areassociated with mineral colloids. In addition, she is investigating theimportance of colloid-assisted transportof actinides in the vadose zone, orunsaturated subsurface, which islocated between the ground surface and the water table. Two-thirds of the underground nuclear tests weredetonated in the vadose zone at theNevada Test Site. The research istaking place in tunnels that have beendug at the site’s Rainier Mesa, whose

vadose zone was previously studied byLivermore scientists.

The colloid discovery, she says,emphasizes the importance of linkingprecisely controlled laboratoryexperiments with field studies. “If you only look at results fromexperiments in the laboratory, youwon’t necessarily understand what’shappening in the field.”

From the arid stretches of theNevada Test Site to physics researchlaboratories of Russia, LawrenceLivermore researchers are pursuingwide-ranging aspects of actinide science.They are combining theory, fieldwork,laboratory experiments, and computersimulations on scales ranging fromatoms to kilometers, all with the aim ofuncovering the secrets of the actinides.

—Arnie Heller

Key Words: actinides, colloids, ForensicScience Center, Glenn T. Seaborg Institutefor Transactinium Science, Nevada TestSite, plutonium, Stanford Linear AcceleratorCenter, Stanford Synchrotron RadiationLaboratory, stockpile stewardship, uranium.

For further information contact Lou Terminello (925) 423-7956([email protected]).

LOUIS J. TERMINELLO is currently leader of the MaterialsScience and Technology Division in the Chemistry and MaterialsScience Directorate at Lawrence Livermore National Laboratory.He is also director of the Glenn T. Seaborg Institute forTransactinium Science at Livermore. He earned his Ph.D. inphysical science from the University of California at Berkeley in 1988 and is an adjunct associate professor there. His research

interests include solid-state physics, atomic and electronic structure determinationof novel materials using synchotron radiation photoemission and absorption, andphotoelectron holography and valence-band imaging studies of electronic materialsurfaces and interfaces.

About the Scientist

Two scanning electron microscope images ofmineral colloids (tiny rock particles) containingplutonium filtered from groundwater at (a) the1-mircrometer scale and (b) the 500-nanometerscale. Mineral colloids consist of clay andzeolites, common secondary minerals found inrocks at the Nevada Test Site. Research by aLawrence Livermore team suggests that smallamounts of plutonium can be transportedconsiderable distances by groundwater byadhering to colloids.

(a)

(b)

500 nanometers

1 micrometer

http://www.llnl.gov/str/Glassley.html

23


Aerogel StructureResearch Highlights

A Predictable Structure for Aerogels

Livermore scientist needs a support system withvirtually no mass for a project he is working on. He is

certain to end up using an aerogel. No mass at all is animpossibility, but aerogels come pretty close. Researchers atLivermore have already synthesized a silica aerogel onlytwice as dense as air.

Sometimes called frozen smoke, aerogels are open-cellpolymers with pores less than 50 nanometers in diameter. In a process known as sol-gel polymerization, simplemolecules called monomers suspended in solution react withone another to form a sol, or collection, of colloidal clusters.The macromolecules become bonded and cross-linked,forming a nearly solid, transparent sol-gel. An aerogel isproduced by carefully drying the sol-gel so that the fragilenetwork does not collapse.

The complicated, cross-linked internal structure givesaerogels the highest internal surface area per gram ofmaterial of any known material. Aerogels also exhibit thebest electrical, thermal, and sound insulation properties ofany known solid.

For about the last 15 years, Livermore has beendeveloping and improving aerogels for national securityapplications. Livermore scientists have also synthesizedelectrically conductive inorganic aerogels for use assupercapacitors and as a water purifier for extracting harmfulcontaminants from industrial waste or for desalinizingseawater. For a time, Livermore was involved with a NASAproject in which an aerogel was to be installed in a satelliteto collect particles of meteorites as they flew by.

Given aerogels’ many sterling qualities, one wouldexpect to find them in use everywhere. Indeed, there hasbeen major industrial interest in aerogels. However, usingthem in everyday applications presents practical problems,specifically the cost of fabrication and processing. Severalyears ago, a Livermore team won an R&D 100 Award fordeveloping a new fabrication method that was faster andcheaper. (See S&TR, December 1995, pp. 22–25.)

A

But another problem still stood in the way. Sol-gelpolymerization is a bulk process with no way to control the size of the sols or the way they come together. Thestructure and density of the final aerogel are dictated tosome extent by the conditions during polymerization such as temperature, pH, type of catalyst, and so on. But withcurrent fabrication methods, the aerogel’s structure cannotbe controlled at the molecular level.

Chemist Glenn Fox is leading a project at Livermore thataims to bring more control to the design and synthesis of

A sol-gel polymer (a) in the sol-gel stage and (b) after it has been dried.This particular sol-gel polymer uses dendrimers and is called a dendrigel.

(a) (b)

http://www.llnl.gov/str/11.95.html

24 S&TR June 2000


organic aerogels. “Laboratory programs would find manymore uses for aerogels if only we could fabricate them toprecise specifications,” Fox says. “They could be used assensors for biological agents, in environmental remediation,as catalysts for chemical reactions, or in experiments on the

National Ignition Facility. Aerogels have also been ofinterest for insulating appliances and homes and for aplethora of other uses. Nanostructured materials areattracting increased scientific and practical interest. Butcontrol of the material’s structure all the way down to the

Aerogel Structure

OH

OHNH2

NH2H2N

N

RR

R

R

N

N

R

R

R

R

R

R

R

R

(a) Traditional organic aerogels start out as either resorcinol or melamine combined with formaldehyde. There is no way to control how the clusterformation takes place, and the end result is a cross-linked polymer resembling a string of pearls. In (b), with dendredic polymers, the design andsynthesis of reactive, multifunctional monomers can be tailored, with specific sites on the molecules activated for cross-linking. The formation andproperties of the resulting gel can thus be carefully controlled.

Generation 1

Generation 2

Generation 3

(a) (b)

25


Aerogel StructureS&TR June 2000

molecular level is needed first.” Fox and a small teamobtained funding from the Laboratory Directed Researchand Development program to apply a relatively newpolymerization method to this problem.

Starting with a TreeDendrimers are highly branched, treelike

macromolecules that can be synthesized “generationally”to produce perfectly regular structures (dendron is theGreek word for tree). Conventional polymers are chains ofdiffering lengths with a range of molecular weights andsizes, while dendrimers have a precise molecular size andweight. Large, multigenerational dendrimers tend to formtidy spherical shapes with a well-defined structure thatmakes them particularly strong.

Fox’s team has begun applying dendritic methodologyto the creation of sol-gels and aerogels in the hope ofachieving structural control. The Livermore team is oneof the first to use dendritic technology in the organic sol-gel process.

Says Fox, “We are trying to understand and controlthe sol-gel polymerization process on a molecular level.Using dendrimers allows us to separate the clusteringand gelling processes when an aerogel is being formed,something that has not been possible before. If wesucceed, the payoff for Laboratory programs will beextremely important. We may be able to script thephysical properties of the aerogel or build specific tagson molecules in a uniform way.”

Organic aerogels are currently formed by combiningeither resorcinol (1,3-dihydroxybenzene) or melamine(2,4,6-triaminotriazine) with formaldehyde. Fox’s team issynthesizing and experimenting with a whole collectionof new starting materials that are being assembled intodendrimers. Some are based on resorcinol to take advantageof its well-documented reactive attributes. Another set ofnew dendrimer systems with rigid cores could give theresulting aerogel greater structural efficiency, improvingthe ease of processing and lowering the cost of aerogelproduction. Other experiments involve the synthesis ofnew organometallic materials and ways to evenly dispersemetal ions in an organic aerogel.

These tailored dendritic monomers are being combinedwith preformed, dendritic, sol-gel clusters whose outer surfacehas been coated to react with the monomer. Two kinds ofdendrimer precursors have been studied, amino-based andaromatic-based, each having different advantages. Amino-based dendrimers are available commercially and have beenstudied extensively. Reactants can be added relatively easily

to their outer surfaces to “functionalize” them, prompting themto cross-link as desired. Benzyl ether dendrimers, on the otherhand, are structurally similar to the colloidal sols of theresorcinol–formaldehyde mix. They are not commerciallyavailable but can be prepared readily in the laboratory.

Controlling the size and composition of the clusters formedduring gelation as well as the type of cross-linking involvedshould give Fox’s team a new-found architectural control overaerogels. Analysis of the structures with infrared spectroscopy,nuclear magnetic resonance spectroscopy, and massspectroscopy will provide a better understanding of howchemistry can affect the composition and structural efficiencyof these nanostructured materials.

—Katie Walter

Key Words: aerogels, dendrimers, polymers.

For further information contact Glenn Fox (925) 422-0455([email protected]).

26 Research Highlights S&TR June 2000


TibetWhere Continents Collide

HE Himalayas get their height from India, and wearen’t talking about genes. About 50 million years ago,

the Indian subcontinent collided with Asia, and the twocontinents continue to converge at a rate of about 5 centimetersevery year. The ongoing collision has been violent enoughto push up the Himalayas, shove Southeast Asia further andfurther southeast, and perhaps most impressively, raise theTibetan Plateau—a landmass as large as two-thirds of thelower 48 states—to an average elevation of 5,000 meters.Uplift of the Tibetan Plateau has been linked tointensification of the Asian monsoon and, by virtue of itserosion products, to gradual changes in seawater chemistry

(a) In this plasticine model, thenorthward movement of a bodyrepresenting India simulates thecreation of faults that haveallowed the southeastwarddisplacement of the SoutheastAsian landmass, among otherfeatures. (b) To accommodatethe extrusion of Indochina, theSouth China Sea opened up tothe east. Continued continentalcollision results in successivelyyounger faults to the north. PaulTapponnier and Gilles Peltzer ofthe Institut de Physique duGlobe de Paris created thismodel. They collaborateregularly with Livermorescientists.

T over long time periods. The Indo-Asian collision thusprovides not only a natural laboratory for studying themechanical response of Earth to plate tectonic forces butalso an opportunity to explore the links between tectonics,climate, and ocean history.

Livermore geophysicists Rick Ryerson, Jerome van derWoerd, Bob Finkel, and Marc Caffee, along with collaboratorsfrom the University of California at Los Angeles and fromParis and Beijing, have been studying this terrestrialwrestling match for several years, making the first-evermeasurements of long-term movement along large faults innorthern Tibet. The Kunlun, Altyn Tagh, and Haiyuan faults

are strike-slip faults that allowblocks of Earth’s crust to slide pastone another, often with disastrousconsequences. All of these faultshave experienced large earthquakesranging in intensity from 7.5 to 8.7.

The function of the faults is asubject of considerable geophysicalcontroversy. Faults may definemajor discontinuities in Earth’slithosphere (the outer 100 kilometersof the crust that define the plates in plate tectonics) and thus absorb a significant portion of theconvergence between India andAsia. Or they may be shallowfeatures that play a secondary rolein a more fluid lithosphere. Someresearch indicates that the Kunlunand Altyn Tagh faults extend tothe base of Earth’s lithosphere,suggesting that they indeed definecontinental plates. A first step indeciding the faults’ extent andfunction is to obtain accurate, long-term slip rates at enough sites alongthe faults to characterize their large-scale behavior.

(a)

(b)

27Continental CollisionS&TR June 2000


AMS and the Dating Game To derive rates of motion along faults, scientists first

identify a site where lateral offset has occurred and thenmeasure the offset and determine its age. Commerciallyavailable satellite imagery, with resolution to 10 meters,allowed the team to select regions where tectonic offsets arebest preserved, such as abandoned stream beds and surfacesformed by glacial action. As shown in the figure on pp. 28–29,the team took measurements at several sites where the faultscross alluvial fans formed by melting glaciers. As a glaciershrinks, the stream running from it becomes narrower,

leaving behind an older, wider streambed in a series ofterraces. The boundaries between different terrace levelsrepresent lateral offset markers.

To determine the age of these surfaces and thus a sliprate, the team relied on experts from Livermore’s Centerfor Accelerator Mass Spectrometry. Conventional massspectrometry measures the concentrations of differentisotopes of the same element. Accelerator mass spectrometry(AMS) does the same job but is much more sensitive thanconventional mass spectrometry. The most common datingmethod is to measure carbon-14 relative to other carbon

Map of Tibet showing the major faults and geopolitical boundaries for reference.

Latitude

Long

itude

28 Continental Collision S&TR June 2000

(a) Site 1 on the Altyn Tagh Fault. The fault is illuminated from thesouth and may be seen as a bright line running across the image. Theactive streambed is on the left and the older streambed terraces risesequentially to the right. (b) and (c) Alternative interpretations of theevolution of the various terraces. (ka = 1,000 years)

Site 1

(a) (b)

(c)


isotopes. AMS, for instance, can find one atom of carbon-14in a quadrillion other carbon atoms, which means thatextremely small samples can be studied.

However, in high, arid mountain ranges, fossil organicremains are often hard to find. Moreover, the ages of somesurfaces may be too old to measure by carbon-14 methods.Therefore, for its first study, on the Kunlun Fault, the teamcompared slip rates obtained through radiocarbon datingwith those obtained by measuring the cosmogenic nuclidesberyllium-10 and aluminum-26 in quartz rock. Cosmogenicnuclides are produced through the interaction of surfacesamples with cosmic rays. Whereas carbon-14 levels fallover time, the levels of cosmogenic nuclides rise the longera sample resides at Earth’s surface. But the amounts theycontain are still small. It has only been in the last 10 years,with advancements in such techniques as AMS, thatmeasuring cosmogenic isotopes has been possible at all.

Under optimal conditions, samples taken from the surfacewill yield the true age of the surface. But a surface may alsocontain samples that were previously exposed to cosmic rays.Or rocks may simply roll downhill and contaminate a previouslyabandoned surface. Scientists must, therefore, collect manysamples, both buried and from the surface, to account for all sources of contamination and derive a site’s true age.

For the Kunlun studies, slip rates derived fromberyllium-10 and aluminum-26 ages compared extremelywell with those from radiocarbon dating.

What the Data SayThe figure at the bottom of p. 28, showing one alluvial

fan along the Karakax Valley segment of the Altyn Tagh Fault,is an example of the many sites studied on both faults. The

fault is clearly visible in part (a) of the figure as is lateraloffset of terrace levels along the fault. Parts (b) and (c) showtwo interpretations of the evolution of this site.

Measurements at 10 sites along the Altyn Tagh Faultyielded slip rates as high as 3 centimeters per year in thewest, decreasing to rates below 1 centimeter per year at thefault’s eastern end. The rate decreases as lateral movementin the west is transformed into the vertical uplift that hascreated young mountains in northeastern Tibet. In contrast,using on measurements at six sites along a 600-kilometerlength of the Kunlun Fault, the team found a uniform sliprate of about 12 millimeters per year over a time span of40,000 years.

Comparison of the two faults suggests that the KunlunFault may be a more mature version of the Altyn Tagh. Whilethe Altyn Tagh Fault is still in the process of propagatingeastward, piling up new mountains along its bow, the Kunlun’smovement appears to be fully transferred to other faults to theeast. But Ryerson is quick to add, “We don’t really understandyet what is happening at the eastern end of the Kunlun Fault.”

These data indicate that the birth and growth of strike-slipfaults has been moving north with time, suggesting that thenorthern portion of the Tibetan Plateau has been uplifted bysuccessive episodes of eastward fault propagation coupledwith the uplift of young mountain ranges. Sediments fromthe young mountain ranges accumulate in closed basins thatare in turn uplifted by fault movement. Ironically, much ofone of the greatest mountain ranges on Earth, the TibetanPlateau, may have been built not of mountains but of basins.

The relatively high slip rates observed along the Altyn Taghand Kunlun faults are consistent with the models showingthat these faults play an important role in accommodatingIndo-Asian convergence. Livermore’s data indicate that themodels representing the lithosphere as a fluid may be flawed.

The first stage of this work, assessing the slip rates onactive faults in northern Tibet, is nearing completion.Barely begun, however, is the next stage—extrapolatingthese observations of active faulting back to the earlyhistory of the collision. Meanwhile, continents continue tocollide in Tibet.

—Katie Walter

Key Words: accelerator mass spectrometry, cosmogenic isotopes,dating techniques, faults, plate tectonics, Tibet.

For further information contact Rick Ryerson (925) 422-6170([email protected]).

29


Continental CollisionS&TR June 2000

2

5 kilometers

Site 3

Three sites studied along the Altyn Tagh Fault are shown. The fault isvisible where it crosses several alluvial fans.


30Each month in this space we report on the patents issued to and/orthe awards received by Laboratory employees. Our goal is toshowcase the distinguished scientific and technical achievements ofour employees as well as to indicate the scale and scope of thework done at the Laboratory.

Patents and Awards

Patent issued to

Peter A. KrulevitchAbraham P. LeeM. Allen NorthrupWilliam J. Benett

Roger D. AinesKent S. UdellCarol J. BrutonCharles R. Carrigan

Layton C. Hale

Patent title, number, and date of issue

Microfabricated Instrument for TissueBiopsy and Analysis

U.S. Patent 5,985,217November 16, 1999

Chemical Tailoring of Steam toRemediate Underground Mixed-Waste Contaminants


Precision Tip-Tilt-Piston ActuatorThat Provides Exact Constraint


Summary of disclosure

A microfabricated biopsy–histology instrument that has severaladvantages over conventional instruments. These include minimalspecimen handling, cutting edges providing atomic sharpness forslicing thin (2-micrometer or less) specimens, use of microlitervolumes of chemicals for treating specimens, low cost, disposability, afabrication process that renders sterile parts, and easy use. The cutter isa “cheese-grater” design comprising a substrate block of silicon that isanisotropically etched to form extremely sharp and precise cuttingedges. Tissue specimens pass through the silicon cutter and lie flat on apiece of glass bonded to the cutter. Microchannels are etched into theglass or silicon substrates to deliver small volumes of chemicals fortreating the specimens. After treatment, specimens can be examinedthrough the glass substrate. For automation purposes, microvalves andmicropumps may be incorporated. Also, specimens in parallel may becut and treated with identical or varied chemicals. The instrument isdisposable because of its low cost and thus could replace currentexpensive microtome and histology equipment.

A method to remediate mixed-waste underground contamination suchas organic liquids, metals, and radionuclides. It involves chemicaltailoring of steam for underground injection. Gases or chemicals areadded to a high-pressure steam flow being injected into wells, towardcontaminated soil located beyond excavation depths. The additives inthe injected steam mobilize contaminants as the steam pushes thewaste through the ground toward an extraction well havingsubatmospheric pressure (vacuum). The steam and mobilizedcontaminants are drawn in a substantially horizontal direction to theextraction well and withdrawn to a treatment point above ground. Theheat and boiling action of the steam front enhance the mobilizingeffects of the chemical or gas additives. While being used to removeany organic contaminants, the method may also be used forimmobilizing metals. An additive can be used to cause metals toprecipitate into large clusters, thereby limiting their future migration.

A device that can precisely actuate three degrees of freedom(commonly referred to as tip, tilt, and piston) of an optic mount. Thedevice consists of three identical flexure mechanisms, an optic mountto be supported and positioned, a structure that supports the flexuremechanisms, and three commercially available linear actuators. Eachflexure mechanism constrains two degrees of freedom in the plane ofthe mechanisms, and one direction is actuated. All other degrees offreedom are free to move within the range of flexure mechanisms.Typically, three flexure mechanisms are equally spaced in angle aboutthe optic mount and arranged so that each actuated degree of freedomis perpendicular to the plane formed by the optic mount. Thisarrangement exactly constrains the optic mount and allows arbitraryactuated movement of the plane within the range of the flexuremechanisms. Each flexure mechanism provides a mechanicaladvantage, typically on the order of 5:1, between the commerciallyavailable actuator and the functional point on the optic mount. Thisfeature improves resolution by the same ratio and stiffness by thesquare of the ratio.

Patents


31S&TR June 2000 Patents and Awards

Patent issued to

Ronald G. Musket

Paul G. CareyPatrick M. Smith

Harley M. Buettner

Matthias FrankCarl A. MearsSimon E. LabovW. Henry Benner


Sharpening of Field Emitter TipsUsing High-Energy Ions


Method of Fabrication of DisplayPixels Driven by Silicon Thin-FilmTransistors


Electrical Heating of Soils Using High-Efficiency Electrode Patternsand Power Phases


Ultrahigh-Mass Mass Spectrometrywith Charge Discrimination UsingCryogenic Detectors



A process for sharpening arrays of field emitter tips of field-emission cathodessuch as those found in field-emission, flat-panel video displays. The process usessputtering of high-energy (more than 30-kiloelectronvolt) ions incident along ornear the longitudinal axis of the field emitter to sharpen the emitter with a taperfrom the tip, or top, of the emitter down to its shank. The process is particularlyapplicable to sharpening tips of emitters having cylindrical or similar(pyramidal, for example) symmetry. The process will sharpen tips down to radiiof less than 12 nanometers with an included angle of about 20 degrees. Becausethe ions are incident along or near the longitudinal axis of each emitter, the tipsof gated arrays can be sharpened by high-energy ion beams rastered over thearrays using standard ion implantation equipment. While the process isparticularly applicable for sharpening arrays of field emitters in field-emission,flat-panel displays, it can be effectively used in the fabrication of other vacuummicroelectronic devices that rely on field emission of electrons.

A method of fabricating display pixels driven by silicon thin-film transistors onplastic substrates. The method is useful for active matrix displays such as flat-panel displays. The process for forming the pixels involves a prior method forforming individual silicon thin-film transistors on low-temperature plasticsubstrates, which are generally considered incapable of withstanding sustainedprocessing temperatures greater than about 200°C. The pixel formation processresults in a complete pixel and active-matrix pixel array. A pixel (or pictureelement) in an active-matrix display consists of a silicon thin-film transistor(TFT); a large electrode, which may control a liquid crystal light valve; and anemissive material (such as a light-emitting diode) or some other light-emittingor attenuating material. The pixels can be connected in arrays wherein rows ofpixels contain common gate electrodes and columns of pixels contain commondrain electrodes. The source electrode of each pixel TFT is connected to itspixel electrode and is electrically isolated from every other circuit element inthe pixel array.

Powerline-frequency electrical (joule) heating of soils using a high-efficiencyelectrode configuration and power phase arrangement. The electrodeconfiguration consists of several heating or current injection electrodes aroundthe volume of soil to be heated and a return or extraction electrode(s) locatedinside the volume to be heated. The heating electrodes are all connected to onephase of a multiphase or single-phase power system; the return electrode(s) is(are) connected to the remaining phases of the multiphase power source. Thiselectrode configuration and power-phase arrangement can be used whereverpowerline-frequency soil heating is applicable. It thus has many potential uses,including removal of volatile organic compounds such as gasoline ortrichloroethylene from contaminated areas.

An ultrahigh-mass, time-of-flight mass spectrometer using a cryogenic particledetector as an ion detector with charge-discriminating capabilities. Cryogenicdetectors have the potential for significantly improving the performance andsensitivity of time-of-flight mass spectrometers, and compared to ion multipliers,they exhibit superior sensitivity for high-mass, slow-moving macromolecularions and can be used as “stop” detectors in time-of-flight applications. In addition,their energy-resolving capability can be used to measure the charge state of theions, which is valuable for all time-of-flight applications. Used as an ion detectorin a time-of-flight mass spectrometer for large biomolecules, a cryogenicallycooled niobium–alumina–niobium superconductor–insulator–superconductortunnel junction (STJ) detector operating at 1.3 kelvin has been found to havecharge discrimination capabilities. Because the cryogenic STJ detector respondsto ion energy and does not rely on secondary electron production (as in theconventionally used microchannel plate detectors), the cryogenic detector candetect large molecular ions with a velocity-independent efficiency approaching100 percent.

(continued on p. 32)


32 S&TR June 2000Patents and Awards

Jim Bryan, a retired Laboratory engineer, was recentlyrecognized by Fortune magazine as one of the six “Heroesof U.S. Manufacturing” of 2000 at a ceremony in Chicago,Illinois. Now in their fourth year, the Fortune awards arepresented annually to innovators who have made a notablecontribution to American manufacturing. Bryan is the firstaward recipient from a U.S. national laboratory.

In the 1980s, Bryan reworked an old British inventioncalled a fixed ball bar by adding a telescoping arm to theinstrument. His invention came about, in part, because ofthe need to produce components with extreme precision for the nation’s nuclear weapons. Today, versions ofBryan’s ball bar are used around the world to test machine-tool performance quickly.

Used by hundreds of companies to determine if theirmachine tools are working properly, telescoping ball barsare plugged into a personal computer for tests in which thecomputer analyzes the deviation of a machine tool’smotion from a perfect circle. (Machine tools, such as lathesand milling machines, precisely cut metal to shape.)

During his Laboratory tenure (1955 to 1987), Bryanmade wide-ranging contributions to metrology andprecision machining. Toward the climax of his career when Bryan was head of the Precision Engineering Group,his team designed, built, and operated the largest diamondturning machine in the world.

Two Laboratory scientists have been elected fellowsof the Optical Society of America.

Stephen Payne, associate program leader in the LaserScience and Technology organization, was recognized for“sustained pioneering contributions to the development ofnovel lamp and diode-pumped solid-state laser materials.”Working with Laboratory colleagues, Payne has developedmore than a dozen laser crystals and glasses. A 15-yearLaboratory employee, he received his Ph.D. in chemistry fromPrinceton University. He has 80 refereed journal publications,holds 11 patents, has received four R&D 100 awards, andrecently received the Excellence in Fusion Engineering Awardfrom Fusion Power Associates.

Mike Perry, associate program leader for the Short-PulseLasers, Applications, and Technology Program, wasrecognized for “pioneering contributions to the developmentand use of high-peak-power, ultrashort-pulse lasers” in high-intensity physics research. Areas of investigation include thefast-igniter concept for inertial confinement fusion andmaterials processing applications in industrial machining andhealth care. Perry was also key in the Laboratory’sdevelopment of large-scale diffractive optics for large-areadiffraction gratings used to manipulate laser light.

Perry has been at the Laboratory for 17 years, starting withhis doctoral work for the University of California at Berkeleyin nuclear engineering/quantum electronics. He has contributedto more than 100 scientific papers on materials processing,diffractive and nonlinear optics, and the use of lasers inmedicine. He holds patents in areas such as inertialconfinement fusion, multilayer dielectric diffraction gratings,and ultrashort-pulse laser machining.

Awards

(Continued from p. 31)

Patent issued to

James C. DavidsonJoseph W. Balch


Vacuum Pull-Down Method for anEnhanced Bonding Process

U.S. Patent 6,000,243December 14, 1999


A process for effectively bonding substrates of arbitrary sizes or shapes. Itincorporates vacuum pull-down techniques to ensure uniform surface contactduring bonding. The essence of the process for bonding substrates such asglass, plastic, or alloys, which have a moderate melting point and gradualsoftening-point curve, involves applying an active vacuum source toevacuate interstices between substrates while providing a positive force tohold in contact the parts that are being bonded. The process enablesincreasing temperature during bonding to ensure that the softening point hasbeen reached and small voids are filled and come in contact with theopposing substrate. The process is most effective where at least one of thetwo plates or substrates contains channels or grooves that can be used toapply vacuum between the plates or substrates during the thermal bondingcycle. Also, it is beneficial where there is a vacuum groove or channel nearthe perimeter of the plates or substrates. In both instances, the process ensuresbonding at the perimeter and reduces unbonded regions in the interior.

33

New Day Dawns in SupercomputingASCI White—the 10-teraops supercomputer developed

by IBM for Lawrence Livermore and due to come onlinein the summer of 2000—is the latest in a series of ultrahigh-speed computers built for the Department of Energy’sAccelerated Strategic Computing Initiative (ASCI). A keycomponent of the DOE’s Stockpile Stewardship Program,ASCI has the goal of providing the computational toolsneeded for running full-scale simulations of nuclear weaponsby 2004. Weapon scientists will use these simulations,archived nuclear test data, and nonnuclear experiments toensure the safety, reliability, and performance of U.S. nuclearweapons. ASCI White is the direct descendant of the ASCIBlue Pacific, a 3-teraops machine. Over the past year, BluePacific has successfully run a variety of unprecedentedthree-dimensional simulations, including the first-ever three-dimensional simulation of an exploding nuclear weaponprimary and an award-winning turbulence simulation.Through DOE’s Academic Strategic Program Alliance,universities have also used Blue Pacific to research unclassifiedproblems applicable to ASCI research, including the physicsof supernovas and fire.Contact:Mark Seager (925) 423-3141 ([email protected]).

Uncovering the Secrets of ActinidesThe elements from actinium (element 89) to lawrencium

(element 103) are known as the actinides. All of the actinidesare metals, and all are radioactive. More complete informationon actinides is needed to better assess the nation’s nuclearstockpile, help stem the clandestine proliferation of nuclearweapons, and gain a better understanding of the use andstorage of nuclear fuels such as enriched uranium. A majorresearch focus is on obtaining a better scientific understandingof plutonium, the most complex and perplexing element inthe periodic table. The Lawrence Livermore chapter of theUniversity of California’s Glenn T. Seaborg Institute forTransactinium Science is attempting to attract and train thenext generation of actinide scientists.Contact:Lou Terminello (925) 423-7956 ([email protected])

Abstracts

U.S. Government Printing Office: 2000/583-047-80034

In the remote desert of theDepartment of Energy’s NevadaTest Site, Livermore physicistsand engineers are conductingsubcritical tests—detonating highexplosives and plutonium withoutsustaining a nuclear chain reaction—to help ensure the safety andreliability of the U.S. nuclearweapon stockpile.

Also in July/August• Accelerator mass spectrometry is on its way tobecoming a routine tool for biomedical research.

• Site-specific seismic analyses help prepareUniversity of California campuses for large-magnitude earthquakes.

• Quantum-dot technology yields ever smallerdevices emitting more colors of light for various applications.

Co

mi

ng

N

ex

t

Mo

nt

h

What ShockedPlutonium Can

Tell Us

University of C

aliforniaS

cience & Technology R

eviewLaw

rence Livermore N

ational LaboratoryP.O

. Box 808, L-664

Livermore, C

alifornia 94551Printed on recycled paper.

Nonprofit O

rg.U

. S. P

ostage

PAID

Albuquerque, N

MP

ermit N

o. 853

supercomputing takes another giant step/67531/metadc741360/... · breakthrough in laser glass...

Documents