The most recent impasse in closureprtKeeciings nearly caused a meltdownin Lithuania's relations with Brussels.In the course of the October 2004parliamentary elections, Prime Minis-ter Algirdas Brazauskas announced hewould keep the plant's first reactorworking beyond its closure deadline atthe end of the year. Voters rewardedhim by returning him to the country'smost powerful office.

Only after the European Commis-sion coldly reminded Brazauskas thatthe decommissioninfi was "enshrinedin Lithuania's accession treaty" didthe prime minister retract the state-ment he made weeks before.

Arturas Dainius, the state secretaryat Lithuania's Economy Ministry,which is in charge of plant closure,said that "the elections didn't playthe least significant role" in the gov-ernment's stance. "You know," headded, "all sorts of 'interesting' ideascan pop up from the pt)litical arena."

Yet the conditions that instigatedthe eleventh-hour crisis over closingthe first reactor will be dwarfed bythe potentially catastrophic issuesLithuania wili face as it prepares toclose the second reactor by the end of2008—another theoretically "en-shrined" date. The energy producedby the first reactor was almost allsent abroad, but the final closure willleave Lithuania able to produce only2.5 percent of its current electricaloutput, leaving a massive void in thecountry's energy supply.

With government officials admit-ting they have no definite plan to re-place the supply from the second re-actor, the hoped-for on-time closureseems doubtful. Casual proposalsabound, but precious few officialideas have surfaced on how to use theaid from Brussels. "We'll either haveto become an energy importer orbuild another plant, in which casewe'll have to decide what type of

plant that will be," said Dainius.Only nebulous suggestions have beendiscussed so far.

Lithuania's power grid has yet tobe connected to the rest of the HU,meaning imported electricity wouldhave to come from Russia—an un-popular move in a country sensitiveto the giant bear's long reach. Andthe prospect of bringing a new nucle-ar reactor online in less than fouryears seems dim given the govern-ment's sluggish pace of decision mak-ing. "Sooner or later the reactor isgoing to have to close, so why don'twe make sound plans for its closurenow?" Jasiulionis asked.

In the meantime, even governmentofficials do not sound confident thatthe second reactor will be closed."We'll live, and we'll see," Dainiustold me. <*

Steven PauUkas is a journalist based inVilnius, Lithuania.

L A S E R E N R I C H M E N T

Separation anxietyBy Jack Boureston & Charles D. Ferguson

N NOVEMBPR 2 0 0 4 , THE ENVl-

ronmental group Greenpeaceaccused the Australian gov-ernment of condoning nucle-ar proliferation by support-

ing the work of a laser uraniumenrichment company named SilexSystems Limited. "If any other coun-try, be it Iran, Syria, or Iraq was in-volved in this research it would betaken as a sign of a covert weaponsprogram," a Greenpeace spokesper-son told reporters.

Nations have been developinglaser isotope separation methods to

enrich uranium for years, but mosthave yet to convert research intocommercial success or have aban-doned laser enrichment altogether.The recent accusations and the diffu-sion of laser enrichment technologiesand know-how as part of peacefulnuclear programs nonetheless againraise tbe question: How much of aproliferation risk does laser isotopeseparation present?

Analysts have paid relatively lessserious attention to the use of laserisotope separation (LIS) to enrichuranium than to the spread of gas

centrifuge enrichment and repro-cessing tecbnology. But certain fea-tures of laser enrichment facilitieswould seem to make them ripe forproliferation—they are typicallysmaller, use less energy, are moreeasily concealed, and may one daybe cheaper to operate than both gascentrifuge and diffusion plants. Still,there are formidable obstacles totbeir development.

Some analysts have regarded laserisotope separation as too difficult tomaster by nations lacking highly ad-vanced technical infrastructures.

14 Bulletin of the Atomic Scientists March/April 2005

One exception is StanleyI'rickson, an analyst atLawrence Livermore Na-tional Laboratory, ln anOctober 2001 paper F.r-ickson warned, "As tech-nology advances, thiswill ntJt remain so." Thisobservation proved pro-phetic in August 2002,when the dissident groupNational Council of Re-sistance of Iran an-nounced at a Washington,D.C, press conferencethat Iran had started anLIS program and devel-oped a laser enrichmentfacility at Lashkar Ab'ad.

The Iranian laser re-search program, whichenriched only milligramsof uranium, had surpris-ingly managed to escapedetection by the Interna-tiona! Atomic FlnergyAgency (IAEA). In Febru-ary 2003, IAEA DirectorGeneral Mohamed El-Baradei acknowledgedthat the IAEA would continue hav-ing problems detecting similar "re-search and laboratory activities" inthe future. But ElBaradei hastenedto add that the IAEA's improvedtechnological capabilities wouldmake it "highly unlikely" that anindustrial-scale LIS program wouldgo undetected.

Another hidden research effortusing laser enrichment came to lightin September 2004, when the IAEAexposed South Korean experiments(for more on this, see "South Korea'sNuclear Surprise," January/February2005 Hullctm). In 2000, scientists atthe Laboratory for Quantum Opticsat the Korea Atomic Energy ResearchInstitute (KAERI) separated about0.2 grams of uranium 235, an iso-tope useful in nuclear fuel or weap-ons, and enriched tbem to levels be-tween 10 and 77 percent. While 20percent is the dividing line betweenlow-enriched and highly enriched

Is laser enrichment a proliferation risk? Greenpeace says "yes" at this September 2003protest in Australia.

uranium, enrichment levels close to90 percent are sought for the pur-poses of making weapons. Sufficientamounts of uranium enriched to 77percent could fuel a nuclear bomb.

Scientists need tens of kilogramsof enriched uranium, more than100,000 times the amount enriched,to make a weapon, and analystsdrew clear conclusions about Seoul'sintentions. "If the question is, couldKAERI have enriched a significantamount of uranium using the facilitythey had in that laboratory, I'm high-ly confident the answer is no," JeffreyEerkens, a leading American laser en-richment expert, told NucleonicsWeek (September 9, 2004). Yet, scientists wouldn't need a commercial-scale US plant to enrich enough ura-nium for a single nuclear weapon ifthey had one or two years to do so.

This is perhaps why South Korea'slaser enrichment activities were ofsome concern to the U.S. govern-

ment. In November 2001, Eerkensgave a presentation on laser isotopeseparation techniques at a scientificconference in South Korea. The nextyear, he proposed to the Energy De-partment that he work with aKAERI scientist to investigate laserseparation of zinc isotopes and otherisotopes useful in medical applica-tions. Energy denied the proposal"because it was too close to uraniumenrichment," Eerkens told Nucleon-ics Week and confirmed for us.

Although the IAEA has repri-manded South Korea for not report-ing its uranium enrichment activitiesin a timely fashion as required by itsSafeguards Agreement, the UnitedStates has not expressed serious con-cern. In early November 2004, be-fore the IAEA Board of Governorsmeeting. Secretary of State GolinPowell said, "I'm quite sure that theIAEA will see it as a minor problemwith experimentation."

March/April 2005 Bulletin of the Atomic Scientists 15

How it works

For more than 30 years, enthusiasts have trumpeted the benefits of laserenrichment of uranium. The energy costs are much lower than tradition-

al enrichment techniques; lasers can, in principle, target uranium 235, theisotope useful for fuel and bombs, while leaving uranium 238, the mostprevalent isotope, alone. But because the two isotopes have very similarexcitation frequencies, the laser beams used to selectively excite uranium235 must be finely tuned and have a narrow frequency spread, which istechnically demanding to achieve.

Scientists have investigated two types of laser enrichment techniques:atomic vapor laser isotope separation (AVLIS) and molecular laser isotopeseparation (MLIS). In AVLIS, tunable dye lasers target vaporized uraniummetal and selectively excite and charge uranium 235. (Copper vaporlasers are typically used to pump the dye lasers.) The positively chargeduranium 235 collects on negatively charged plates and is then extractedin a labor-intensive process. In principle, a single-stage AVLIS plant couldmake low-enriched uranium suitable for reactor fuel, and a three-stageplant could produce about onebomb's worth {25 kilograms) ofweapon-grade uranium per year.Inadvertently charging and col-lecting uranium 238 is a majorproblem with this technique.

In MLIS, an infrared laser is di-rected at uranium hexafluoridegas. The laser excites uranium235 hexafluoride gas, while notdisturbing the uranium 238 hex-afluoride gas. Another laserknocks a fluorine atom off an ex-cited uranium 235 hexafluoridemolecule, creating uraniumpentafluoride, which precipitatesout as a white powder and con-tains enriched uranium. Because producing even low-enriched uranium forcommercial purposes is very difficult for a single-stage MLIS plant, cas-cading is necessary. Each stage in the cascade requires refluorination ofthe uranium pentafluoride, a process that can substantially increase thecosts of MLIS when compared to AVLIS.

Jack Boureston & Charles D. Ferguson

The AVLIS setup at Lawrence LivermoreNational Laboratory.

South Korea and Iran were neitherthe first nor the last to experimentwith laser enrichment. In fact, morethan 20 other countries have re-searched laser isotope separationtechniques. They include Argentina,Australia, Brazil, Britain, China,France, Germany, India, Iraq, Israel,Italy, Japan, the Netherlands, Pak-istan, Romania, Russia, South Af-rica, Spain, Sweden, Switzerland, theUnited Stares, and Yugoslavia.

Most of these nations have con-fined their LIS work to the labora-tory. Britain, France, and the Unit-ed States have developed LISprograms that could move beyondthe lab into the pre-industrial phaseand ultimately into commercialproduction. Australia and Japan arealso known to have invested signif-icant resources in trying to buildLIS uranium enrichment plants, butno country currently operates a

commercial-scale laser enrichmentfacility for separating uranium, ac-cording to the Uranium InformationCentre in Australia.

AVLIS has left the buildingThe United States was on the vergeof commercialization, when USEC,then known a.s the U.S. EnrichmentCorporation, decided in June 1999to cancel its atomic vapor laser iso-tope separation (AVLIS) program.This came as a surprise consideringUSFC had spent roughly $100 mil-lion on AVLIS since being privatizeda year earlier. In total, the U.S.AVLIS program involved 27 years

of research and develop-ment and an investment ofsome $2 billion. USECs costestimates to make AVLISready for commercialization,which soared into the hun-dreds of millions of dollars,were a major factor in theprogram's cancellation.

U.S. interest in laser en-richment methods did notimmediately die with theAVLIS program. In 1996,USEC had invested in theseparation of isotopes by the

laser excitation (SILFX) techniquebeing developed in Australia. TheU.S. and Australian governmentsdemonstrated further interest in theproject on June 20, 2001, whenthey announced that they had offi-cially classified the SILEX method.But USECs interest in SILEXproved short-lived, and on April30, 2003 it terminated SILEXfunding.

In a May 1, 2003 press release,Michael Goldsworthy, chief execu-tive officer of Silex Systems, saidthat his company disagreed withUSECs view of the potential andcurrent state of the SILEX technolo-gy. "It is incomprehensible to us thatUSEC has decided to abandon theSILEX program only a few monthsshort of completing the current testprogram and being in a good posi-

16 Bulletin of the Atomic Scientists March/April 2005

tion to assess the economic perfor-niLiiicc of the SIL.EX process/' hesaid. As of July 2004, Silex System.swas srill studying the economicfeasibility of its laser uranium en-richmenr method.

After abandoning laser isotopeseparation methods altogether,USfiC," is now focusing on plans tobuild a centrifuge plant in Piketon,Ohio, the site of the former Ports-mouth Gaseous Diffusion Plant. In2005, USHC hopes to begin operat-ing the American Centrifuge Demon-stration Facility at the site, and by2010, the company wants to con-struct the most efficient centrifugeplant in the world with an annualcapacity of 3.5 million separativework units of enrichment. This plantcould provide enough fuel yearly forhctwcen 30 and JiS 1,000-megawattlight-water reactors.

Gas centrifuge and diffusion meth-ods continue to dominate the globalnuclear fuel enrichment market. Bychoosing to build a centrifuge en-richiiient plant, USEC went with aproven commercial method, reduc-ing financial risks. Although a cen-trifuge facility requires iiiore energyto run than a laser enrichment plant,it will save substantially compared toUSECs remaining gaseous diffusionplant in Paducah, Kentucky., whichis approaching the end of its life.

Laser allureThe potential of LIS has attractedstates eager to establish a domesticenrich inent capability. Like theUnited States, Brazil invested in alaser enrichment program for manyyears hut eventually chose to build acommercial centrifuge plant. Brazil'slaser project aimed to "demonstratethe technical viability of the laserprocesses for isotope separationusing, as long as possible, resourcesavaiiabie in Brazil," according to a199S scientific article hy a researchteam centered at the Instituto de Es-tudos Avani^iidos. That report cau-tioned, "A successful commercial-

ization of this technology willthreaten well-established fuel-cycleactivities."

Brazilian laser research planningoriginated in the United States in theearly 1970s, and was led by SergioPorto, a Brazilian scientist and profes-sor at the California Institute of Tech-nology and the University of SouthernCalifornia who took a special interestin educating Brazilian students. Portoreturned to Brazil in 1975 to head itslaser enrichment program, which wascontrolled by the air force and sup-ported by the National Nuclear Ener-gy Commission.

The Brazilian program has hadmixed results. Brazilian researcherswe contacted declined to release de-tails of how much uranium theyhave enriched and to what level ofenrichment, but Eerkens, the laserenrichment expert, told us that hebelieves they probably attained alevel of 50 percent enrichment. Un-certainty remains about the totalamount of uranium enriched. Andthe program has struggled with ex-traction and separation of the en-riched uranium.

Brazilian scientists take pride inboth their laser and centrifuge ac-complishments, and Brazilian resis-tance to allowing IAEA inspectorsfull access to the centrifuge facilitymade headlines in September andOctober 2004. Eduardo Campos.,Brazil's minister of science and tech-nology, has claimed that the cen-trifuge technology at the facility is"100 percent" Brazilian. Outside ex-perts douht this claim, yet broaden-ing domestic enrichment capabilities,in Brazil and elsewhere, presentsome concern.

Eaced with stiff competition fromgas centrifuge facilities, laser enrich-ment programs may never see large-scale coniiiiercialization. But LISmight offer a clandestine means forproducing a few hombs' worth ofweapon-grade uranium. Accordingto Stanley Erickson's paper, "the re-quirements for a proliferation facili-ty usiiig LiS are not as strenuous as

they are for commercial productionof nuclear fuel."

During the past few decades, thematerials and know-how necessaryto build laser systems useful for ura-nium enrichment have spread wide-ly. Eor instance, copper vapor lasersystems, once costly and technicallyprohibitive, ore now built by highschool students for science projectsand employed in undergraduatelaser laboratory experiments. Dyelasers, another essential componentof AVLIS enrichment, are also he-coming more available. "This diffu-sion of laser knowledge and experi-mental interest means that expertiseabout the finer points of laser con-struction is spreading and will makedevelopment of [lasers| easier," ob-served Erickson.

LIS facilities could escape detec-tion more easily than traditional ura-nium enrichment plants. They wouldbe more compact than a centrifugeplant with an equivalent capacity,and a lot of LIS research has takenplace at universities, which are typi-cally not safeguarded facilities.Moreover, the lasers used in LIShave several dual-use applications.This was one of the concerns raisedabout Silex Systems' operations inAustralia. In response to the Novem-ber Greenpeace report, Australia'sdefense minister, Robert Hill, ac-knowledged that dual-use materialsfrom Australia might have been "in-nocently" exported and used withinan unnamed country's nuclearweapons program.

Another worry is that a nation canhone its LIS skills with elementsother than uranium, while develop-ing a breakout capability and mini-mizing the potential triggering ofsafeguards. To enrich other elementsrequires the same LIS equipmentused to enrich uranium, only tunedto different laser frequencies. Onesuch element is ytterbium, which hasfew industrial applications. "If theyknew how to enrich ytterbium, thenthey could enrich uranium," accord-ing to Eerkens.

March/April 2005 Bulletin of the Atomic Scientists 17

Export controlsNonproliferation analysts have de-hated whether current export con-trols could better address the risk ofLIS proliferation. At a 1999 scienceand security conference at EudanUniversity in Shanghai, DavidDaniels, then a Harvard graduatestudent in physics, concluded, "Thecurrent international system is ableto safeguard a declared LIS facility,but it is not able to timely detect thediversion of significant amf)unts ofuranium to a LIS facility." However,this international system "may pre-vent the construction of new plantsthrough appropriate export con-trols." In contrast, Erickson's paperwarned that advanced technologiescould "outflank" export controls in"nations with moderate-size econo-mies during the first decades of thetwenty-first century."

To combat these possibilities, Er-ickson recommends creating a com-prehensive map of proliferation thatwould require "more labor-intensive"research on the "training, work ex-

perience, and scientific interchange ofnationals of a possible proliferatingnation." The map would also identi-fy the "list of technical achieve-ments" required for each LIS methodand "what scientific knowledge orengineering know-how is needed foreach." Daniels has encouraged a re-examination of the ability of exportcontrols to catch clandestine opera-tion of LIS plants.

Any steps to prevent the spread ofLIS technologies would require agreater technical understanding than isused in establishing conventional ex-port controls. Eirst, officials should de-termine what aspects of LIS can maskas civilian research and assess wheresuch research is taking or may takeplace. While the IAEA's AdditionalProtocol requires nations to declareLIS facilities, identifying the nascentphases of LIS research that could even-tually be applied to uranium enrich-ment presents another challenge.

The nonproliferation regime alsoshould guard against an A. Q. Khan-like theft of LIS uranium enrichment

technology. In the 1970s while work-ing for the uranium enrichment con-sortium URENCO, Khan, a Pak-istani metallurgist, stole designs foruranium centrifuges. He used thispurloined knowledge to huild Pak-istan's centrifuge enrichment plants,which supplied the highly enricheduranium for Pakistan's nuclearweapons. He also distributed the de-signs and centrifuge componentsthrough a nuclear black market.Today, using US for bomb-scale ura-nium enrichment appears out of po-tential proliferators' reach, hut thefurther spread of LIS expertise andtechnologies increases the risk thatsomeday another Khan will peddlethese tools to the highest bidder. ^

Jack Boureston is managing directorand senior research analyst at First-Watch International, a private weap-ons of mass destruction proliferationresearch group in Monterey, Califor-nia. Charles D. Ferguson is a scienceand technology fellow at the Councilon foreign Relations.

I D A H O

Nuclear comes home to roostBy Jonas Siegel

S iNci: 1949, THE ENERGY DH-

partment, other federal agen-cies, and the navy have builtand tested .52 nuclear re-search reactors in the highdesert plains of southeastern

Idaho; a few still operate today.Number 53 might be built at the newIdaho National Laboratory (INL), ifthe Energy Department has its way.

INL, formed on Eebruary 1, is aconsolidation of the Idaho NationalEngineering and Environmental Lab-

oratory (INEEL) and Argonne Na-tional Laboratory-West. "Our goal,within this decade, is to have this labemerge as one of the premier appliedresearch and nuclear engineering in-stitutions in the world," then-EnergySecretary Spencer Abraham said inApril 2003.

The renewed emphasis on nuclearenergy research is one aspect of Ener-gy's expanding Idaho operations.The laboratory will also continueplaying a greater role in homeland

security-related research, and mayeventually house al! of Energy's workon plutonium-decay power genera-tors for spacecraft, including a newplutonium 238 processing facility.But INL faces doubts about its newmissions, as well as questions andcriticisms from Idahoans.

Investing in nuclear enerqyIn 2002, the United States and 10other members of the Generation IV

18 Bulletin of the Atomic Scientists March/Apri! 2005