.*

Zheng-Tian Lu’”,

Atom Trap? Trace Analysis

Kevin Baileyi, Chun-Yen Chen’, XU-DU”3, Yi-Min Lil,Thomas P. OConnor’, Linda Young2

‘Physics Division, ‘Chemistq Division, Argonne National Laboratory, Argonne, IL 60-/39, 0S4‘Department of Physics and Astronomy, North>vestern University, Evanston, IL 60208, US,-t

Abstract. A new method of u!trasensitive trace-isotope analysis has been developed based uponthe technique of laser manipulation of neutral atoms. It has been used to count individual 85Krand *[IQ atoms present in a natural krypton samp[e with isotopic abundances in therangeof 10*]and 10’3, respectively. The atom counts are free of contaminationfrom other isototopes,elements,or molecules. The method is applicable to other trace-isotopes that can be efficientlycaptured with a magneto-optical trap, and has a broad range of potential applications.

INTRODUCTION’

Much can be learned from the concentrations of the ubiquitous long-livedradioactive isotopes. W. Libby and coworkers first demonstrated in 1949 that traceanalysis of lJC (tln = 5715 yr, isotopic abundance = 1x10“*2) can be used forarchaeological dating [1]. Since then, two well established methods, low-levelcounting [2] and accelerator mass spectrometry [3], have been used to analyze manymore trace-isotopes at about the parts-per-trillion level and to extract valuableinformation encoded in the production, transport, and decay processes of theseisotopes. The impact of uhrasensitive trace-isotope analysis has reached a wide rangeof scientific and technological fields. We have recently deveIoped a new method,Atom Trap Trace Analysis (ATTA) [4], which is capable of analyzing trace-isotopeswith an isotopic abundance at the parts-per-trillion level. This new method promisesto enhance the capability and expand the applications of ultrasensitive trace-isotopeanalysis.

In this paper, we will first describe the motivation of analyzing 8‘Kr and 8%.r. Wewill then survey the existing techniques, followed by a description on ATTA and ourwork on the detection of the rare krypton isotopes. Finally, we will discuss some otherpotential applications of ATTA.

“Email:lu@anl.gov URL: www-mep.phy.anl.gov/attti

The submitted manuscript has been crestby the University of Chicago as operatorArgonne National Laboratory (“Argonntunder Contract No. W-31-109-ENG-38 wthe U.S. Department of Energy. The U,Government retains for itself, and others acing on its behalf, a paid-up, nonexclusiveirrevocable worldwide license in said articto reproduce, prepare derivative works, ditribute copies to the public, and perform pu:Iicly and display publicly, by or on behalf tthe Government,

DlSClAlME13

This report was prepared as an account of work sponsoredby an agency of the United States Government. Neitherthe United States Government nor any agency thereof, norany of their employees, make any warranty, express orimplied, or assumes any legal Iiabiiity or responsibility forthe accuracy, completeness, or usefulness of anyinformation, apparatus, product, or process disciosed, orrepresents that its use would not infringe privately ownedrights. Reference herein to any specific commercialproduct, process, or service by trade name, trademark,manufacturer, or otherwise does not necessarily constituteor impiy its endorsement, recommendation, or favoring bythe United States Government or any agency thereof. Theviews and opinions of authors expressed herein do notnecessarily state or refiect those of the United StatesGovernment or any agency thereof.

Portions

DISCLAIMER

of this document may be illegiblein electronic image products. Images areproduced from the best available originaldocument.

*

RARE KRYPTON ISOTOPES

Kqpton gas constitutes 1 ppm of the earth atmosphere in fraction volume. It has.+ stable isotopes, 78Kr(isotopic abundance = 0.35’%0),80Kr(225%), 8%r( 11.6’%0),

8’Kr(1I.5’%). 84Kr(57%), 86Kr(17.31%0),and ttvo long-lived radioactive kotopes, 8’Kr48 IK atoms and 3x 10585Kratoms in 1 literand ‘~Kr (Table 1). There are about 2x 10

STP of air. There are roughly 1038[Kr atoms in one kilogram of modern water or ice.81Kr is produced in the upper atmosphere by cosmic ray induced spallation and

neutron activation of stable k~pton isotopes [5]. Throughout its long lifetime in theatmosphere, 81Kr is }vell mixed and distributed over the earth with a uniform isotopicabundance. Human activities on nuclear fission have had a negligible effect on the8iKr concentration, largely because the stable 81Br shields 81Kr from the neutron richisotopes that are produced in nuclear fission [6]. These physical and chemicalproperties make 81Kran ideal tracer for dating ice and groundwater that are older than100ZOOOyears [7], which is beyond the ran e of lJC-dating.

8’Kr is a fission product of E23’U and 9Pu. ‘ Its present-day concentration in theenvironment has primarily been released by nuclear fiel reprocessing plants. As aresult, its abundance in the atmosphere has increased by six orders of magnitude sincethe 1950s. It has been used as a general tracer to study air and ocean currents [8], dateshallow groundwater, and monitor nuclear-fuel reprocessing activities [9]. Due to itsfast mobility, it may b.e used as a leak sensor to check the seals of nuclear fiel cellsand nuclear waste containers.

Noble gas tracers in general have the advantages that they can be chemicallyseparated from large amount of raw samples, and that their transport processes inenvironment are easier to understand.

Table 1. The properties of Kr-81 and Kr-85.Isotope Half-life Atmospheric Isotopic Applications

(year) abundanceIQ-8 I 2.3X1O’ (5.9t0.6)x10-1’, LLC[5] Geologicaldatingofpolariceandgroundwater

(4.5f0.3)x10’3, LLC[10](5.3*1.2)X10-’3,AMS[6](1.0k0.4)x10’2,ATTA[4].—., ....—... ——___________

IQ-85 10.8 -1 Oyl;%~C[8]—.——

Morutornuclear-fuelre~ocessingactivities

EXISTING TECHNIQUES.

Here \ve briefly review the existing techniques with an emphasis on the analysis ofthe rare krypton isotopes.

Low-Level Decay Counting (LLC)

Long-lived trace-isotopes decay in vairous ways including a-decay, ~-decay, or e-capture, A single decay event releases energy in the range of 10J-I O’ eV, and can bereadily detected with a scintillation or a proportional counter with high efficiency

,

(>50’Ho).The o~erall detection etliciency is usualiY limi[ed by the short counting time(t,) compared \vith the half-life of isotopes (tl,t) since the fraction of nuclei decayedduring the counting time.~= ln2x(tC/tl,2)\vhen t, << tl ~. For example, in one ~veekofcounting. 10-3of ‘>Kr, 10-6of lJC, or 10”7of ‘iKr uould decay. The shorter the half-

I life, the more efficient this method is.LLC is often carried out in a specially designed underground laboratory in order to

avoid background due to cosmic-rays and the radioactivity present in commonmaterials. Environmental samples often contain other radioactive isotopes, which canbe reduced by chemical purification or, in the case of short-lived impurities such as%1 (tl,~= 3.8 days), by \vait@g.

LLC is used to analyze *’IQ [8]. It \vas also used in the first observation ofatmospheric- 81Kr [5]: but can no longer be used because in today’s air the decayactivity of 8>Kr is 10> times that of 8lKr. Pre-nuclear-age samples are also affected,and are now extremely difficult to analyze with LLC due to the inevitable smallcontamination of modem krypton during the sampling and preparation stages.

lh!lass spectrometry (MS)

Atom counting has a number of advantages over decay counting. The efficiencyand speed of atom counting is not fbndarnentally limited by the long half-lives ofisotopes, nor is it affected by radioactive backgrounds in the environment or insamples.

The most popular atom-counting method is mass spectrometry, which can separateand detect individual ions of a chosen mass. However, the sensitivity of this method islimited by intetierence from isobars, i.e., atoms of other elements or moIecules withthe same mass number, w-hichare present in trace amount even after the most carefulchemical purification. Mass spectromet~ in general has a detection limit at an isotopicabundance level of 10-9[11], and is not suitable for analyses at the PPT level.

Accelerator mass spectrometry (ANIS)

Isobar contamination can be eliminated in some cases by performing massspectrometry with a high energy (-MeV) beam from an accelerator [12, 13, 14]. First,molecular isobars can be eliminated by passing the accelerated beam through a thinfoil where molecules disintegrate. Second, some isobar atoms can be eliminated byexploiting the stability property of negative ions that are used in the first accelerationstage of a tandem accelerator. For exampIe, %J-, the only abundant isobar of lJC-, isnot stable, and consequently not accelerated.

The advantages of atom counting are indeed realized with AlMS, which hasreplaced LLC as the standard method of lJC-dating. Furthermore, AMS has openedup new applications by analyzing other long-lived trace-isotopes that were reviousIy

fimpossible to do. The most utilized isotopes in AMS are 10Be, lJC, *sAl, ‘Cl, ‘lCa,and 1Z91[3]. The AMS community has grown steadily since the late 1970s, even at acost of severs! million U.S. dollars for a typical AMS setup. As of 1998, there were

about 40 dedicated A\lS facilities around the \vorld. and more facilities $vhere AMS\\orks are carried out [15].

K~pton isotopes can not be analyzed at a standard .4MS faciiit~ that uses a tandemaccelerator because they do not form negative ions. A new approach has beendeveloped by P. Collon et. al. [6]. \vho used an ECR source to produce positive@pton ions, and a high energy (GeV) cyclotron (Kl 200 Superconducting Cyclotron,hlSU) to produce a fully stripped krypton ion beam. Once fully stripped, 81Kr+3Gcanbe cleanly separated from its abundant isobar 8‘Br (2=35). They have thus realizedradio~pton dating and determined the ages of ground~vater, ranging from 200 to 400kyr, at several sites in the Great Artesian Basin in Australia [15]. In a typical run, 16tons of ground~vater w-ere processed to extract 0.4 cm3 STP of krypton gas, andresulted in 60-100 81Kr counts with a detection efficiency of-1x10-5. Because old iceis much more difficult to extract than old groundwater in similar quantities, animproved efficiency, at 10-3or higher, is needed before dating ice can be realized.

Resonance ionization mass spectrometry (RMMS)

An alternative method of reducing isobar contamination is to use. resonant photonsto selectively ionize the element of choice [16, 17]. Figure 1a shows a two-stepionization scheme where the first step is a resonant excitation, followed by a secondnon-resonant ionization step [18]. More steps of resonant excitation can be added toenhance the selectivity. The combination of isobar selection by resonance ionizationand isotope selection by mass spectrometry would, in principle, enable RIMS to reacha selectivity well below the PPT level. In practice, however, complications such asthermal or collisional ionization may limit the detection sensitivity and selectivity.

G.S. Hurst and coworkers have counted 811Qatoms using the excitation schemeillustrated in figure 1a [18]. Krypton is a difficuh case because the first excitationneeds a laser of 116.5 nm wavelength. In their work, 81Krhad to be pre-en.riched witha mass spectrometer three times to reach an isotopic abundance of -104 beforecounting was performed. The whole process, including the enrichment cycles, has atotal detection efficiency of >50°/0. This scheme, although extremely efficient,involves a multi-step operation that is difficult to implement in practical applications[19].

RIMS is successfully implemented in cases }vhere cw narrow-bandwidth lasers areavailable to excite atoms. For example, K. Zimmer et. al. have analyzed 90Sr, anuclear fission product, in dust particles collected in Munich after the ChemobyIaccident [20]. Applying collinear laser spectroscopy on a fast atomic beam, they havereached an isotopic selectivity (Sr/90Sr)of> 1011w-ith a detection limit of 3x 107 atoms.A simplified version [21], in which laser spectroscopy on a thermaI atomic is applied,has demonstrated an isotopic selectivity (Sr/90Sr) of >1010 with a detection limit of5x 10s atoms. Work aimed at analyzing ‘lCa at a higher selectivity is in progress [22].-- --

Photon-burst mass spectrometry (PBNIS)

A single atom can also be detected by obsewing its fluorescence burst in a resonantlaser beam [23]. By detecting multiple photons in coincidence during the short transittime, both the detector dark counts and the noise photon counts due to light scatteredoff \valls can be suppressed. Furthermore, multiple photon detection also enhancesisotopic selectivity.

Using PBMS, W. N1. Fairbmk jr. and co~vorkers detected 85Kr at the isotopicabundance of 6X10-9 [24]. In their \vork, metastable krypton atoms in a fast beam,produced by neutralizing a mass-selected ion beam, are counted when passing througha photon-burst detection region that consists of ten avalanche photodiode detectors.They pointed out that, by doubling the number of photon detectors and therebyimproving the photon collection efficiency, this method may detect 8’Kr at theatmospheric abundance level (-101 l).

13.D.Cannon et. al. propos~d a new concept [25], in which a resonant laser beam isused to transversely deflect 8’Kr atoms out of the primary metastable krypton. Theseparated 85Kr atoms are then detected with the photon-burst technique. Theysucceeded in enriching 8?Q in the deflected beam by a factor of 1.2x 10~, wlich islimited by non-resonant deflection due to collisions [26].

ATOM TRAP TRACE ANALYSIS (ATTA)

ATTA is a new laser-based atom-counting method [4]. It has been used to analyzeboth 81Krand 8’Kr in an atmospheric krypton sample with no other isotope enrichmentprocesses. The atom counts contain no contamination from other isotopes, elements,or molecules. The isotopic selectivity (Kr/81Kr) has reached 1x1013, and is currentlylimited by the operation time. ATTA can tolerate impure gas samples, and does notrequire a special operation environment.

Our design is based on a type of magneto-optical trap system that had been used totrap various metastable noble gas atoms [27, 28]. Trapping krypton atoms in the5s[~/z]z metastable level (lifetime = 40 see) is accomplished by exciting the 5s[3/2]2-5p[’/z]J transition (Fig. lb). Two repump sidebands are generated via additionalAOIMSto optically pump the atoms into the F=l 3/2 level for 8’Kr and F=l 1/2 level for8*Kr where they can be excited by the trapping light. In the analysis, a krypton gassample is injected into the system through a discharge region, w-here about 1x104 ofthe atoms are excited into the 5s[J/2]z level via electron impact excitation. Two-dimensional transverse cooling is used to reduce the atomic beam divergence andamplify the atom flux in the forward direction by a factor of 20. The thermal (300”C)atoms are then decelerated with the Zeeman slowing technique [29, 30], and loadedinto a MOT [31]. Atoms remain trapped for an average of 1.8 sec as the vacuum ismaintained at 2X10-8Torr. This trap system can capture the abundant 8~Kr atoms atthe rate of 2xIOS see-l. The ratio of the capture rate to the injection rate gives a totalcapture efficiency of1x10-7.

.

\\’ith expected capture rates bet~~een 10“3Scc”i and 10-LSeC-}for the rare k~ptonisotopes, the system must be capable of detecting a single atom in the trap [32, 33]. Inthe trap. a single atom scatters resonant photons at a rate of -107 see-l, of \vhich 1%are collected. spatially filtered to reduce background light, and then focused onto anavalanche photodiode \vith a photon counting efficiency of 23V0. In order to achieve ahigh capture efficiency and a clean single atom signal, the setup is sl~itched at 2 HZ

between the parameters optimized for trap capture and for atom counting. Theresulting fluorescence signal of a single atom is 16 kcps (kilo-counts per second) whilethe background level is 3.4 kcps (FiU. 2).

‘Kr and ‘1Kr atoms from a natural krypton gasWe have trapped and countedsample. The frequency settings of the trapping laser and the tlvo sidebands are in

food agreement with previous spectroscopic measurements obtained using enrichedjKr gas and enriched 81Kr gas [34]. We have also mapped the atom capture rate

versus laser frequent y (Fig. 3). Furthermore, repeated tests were performed underconditions in \vhich a 8% (8*IQ) trap should not work, such as turning off repumpsidebands and tuning the laser frequency above resonance, and these tests alwaysyielded zero atom counts. These tests show that the recorded counts are solely due tolaser-trapped 85Kr(8*Kr) atoms.

Previous efforts to develop a laser-based technique have encountered seriousproblems with contamination from nearby abundant isotopes. ATTA is immune fromisotope contamination, for several reasons: fluorescence is only collected in a smallregion ($0.5 mm) around the trap center; a trapped atom is cooled to a speed below 1m/s so that its laser induced fluorescence is virtually Doppler-free; the longobservation time (>100 ms) allows the atom to be unambiguously identified (S/N =40); and trapping allows the temporal separation of capture and detection so that bothcapture efficiency and detection sensitivity can be optimized. Our design also providesadditional features, such as chopping off the atomic beam before detecting the trappedatom or sending in a laser beam to selectively push out (or de-excite) the atoms ofcontaminant isotopes.

The capture rate of our system depends on the discharge current, laser power, as\vell as optical alignment, At one particular setting, we measured capture rates of%r, 8?Cr, and 8[Kr, which were (1.5+0.3)x108 see-l, (1.9*0 .3)x10-z see-*, and(1.3~0.4)x 10-3see-l respectively. If we assume the same detection efficiency for allthree isotopes, then we get isotopic abundances of (1.5+0.4)x10-11 for 85Kr and( 1.O+O.4)X10-*Zfor 81Kr, which are in good agreement with previous measurementsusing other methods [5, 6, 10]. The capture efficiencies can be calibrated withenriched samples of known isotopic abundance to correct for any isotope-dependenteffects and measure isotopic ratios in unknown samples. For example, in 81Kr-dating,a know-n amount of 85Kr can be mixed into the sa&ple, thus allowing the 81IQabundance be extracted by measuring the ratio of 8*Kr/85Kr.

Efficiency

Our system has achieved an efficiency of 1x 10”7. Use of this system to measure theabundance of ‘~Kr to }vithin 10“/O~vould require 2 hours and a @pton sample of 3 cm3STP ~~hile measurement of ‘lKr to \vithin 10% Yvouldrequire 2 days and a sample of60 cm3 STP. This limits the current system to atmospheric applications where largesamples of gas are available. Improvements, such as a liquid-nitrogen cooleddischarge source and recirculation of krypton gas [35]. are presently underinvestigation.

The metastable level of krypton can also be populated via photon excitations (Fig.1c, Id). With a suitable laser or lamp, the excitation efficiency could be much higherthan the - 1x104 currently achieved with a discharge. Furthermore, without theconstraint on gas pressure imposed by a discharge, atoms can be w-elIcollimated andcooled, thus further reducing the inefficiency due to beam divergence.

Trace AnaIysis of Cesiuin Isotopes

By combining arecently analyzed

mass separator and a MOT, Dave Vieira and coworkers havetwo Long-lived radioactive isotopes, 135CS and 137CS [36].

Applications of this analysis ‘are in environmental science [37] and nonproliferationmonitoring. In this }vork, cesiurn is ionized, accelerated, mass selected, and implantedinto a foil located inside a glass celI. Neutral atoms released from the foil are thencaptured by a MC)T via the vapor cell loading technique, and detected by observingtheir fluorescence in the trap.

With a particular sample, where the isotopic abundances of 13jCs and 13YCSareabout 10-6-10-5,the MOT holds -20,000 atoms. By comparing the tra fluorescenceof both 135CSand f-‘37CS. They have determined the abundance ratio 1 ‘Cs / 137CS=1.21 k O.lostatk().30SYS.

OTHER POTENTIAL APPLICATIONS

Laser manipulation of neutral atoms has been demonstrated on alkali metal (Li, Na,K, Rb, Cs, Fr), alkali earth (Mg, Ca, Sr), noble gas (He, Ne, Ar, Kr, Xc), andindividual cases including Al, Cr, Ag, and Yb. Based on these demonstrated cases,there are already 16 long-lived radioisotopes (Fig. 4) that can be analyzed with ATTAfor a broad range of applications. The conference proceedings of ALMSand RIMS aregood references where the applications are discussed. Some examples are discussedhere.

Solar neutrinos

Solar neutrinos were first detected with a radioisotope tracer [38]. In thisexperiment, 37Ar (t112= 35 days) were produced at the rate of 0.5 atoms/day in a

sealed tank containing 6 i5 tons of perchlorethylenc via the ~TCl(vt. e)~7Ar(threshold =0.814 hleV) reaction. These 37Aratoms \vere recovered \vith over 900/0 efficiency andcounted in a proportional counter. LLC \vorks very ~~ellhere due to the short half-life,and that the source material is artificially maintained in a clean environment. Themeasured neutrino flux disagrees with the prediction of the Standard Solar Model [39],indeed the measured ‘B and ‘Be neutrino flux is 2.56t0.22 SNU \vhile the theoreticalprediction is 7.7~1.2 SNU, thus establishing the so-called solar neutrino puzzle. Thelater experiments as \vell as theoretical \vorks continue to support this disagreement[40].

An analogous detector that counts 81Krproduced in the 8]Br(v., e)81Kr(threshold=0.471 MeV) reaction has been proposed [41, 39]. Due to a lower reaction threshold,this detector \vould be more sensitive to the 9Be (3 10/0)neutrinos than the chlorinedetector (15%). For a detector containing 1000 tons of a Br-chemical, 81Kr isproduced at a few atoms per day. Thus an atom-counter of -10% efficiency isrequired.

Besides of measuring the present-day solar neutrino flux, radioisotopes can play aunique role in measuring the long-term average of the solar neutrino flux, and testingthe long-term thermal stability of the solar core [42, 43]. In such a geochemicalexperiment, tracer atoms (for example, 98Tc)are chemically extracted from an ancientmineral deposit (for example, moIybdenite), and then counted. The number of traceratoms present in the target mineraI is a measure of the solar neutrino flux integratedover the lifetime of the tracer. Suitable geological deposits have been located for the98Tcand 203Pbexperiments. However, an efficient (10-3)and unambiguous method ofcounting these tracer atoms has to be developed before such experiments can becarried out. MS [44], AMS [45], and REMS [46] have been attempted, but none hassucceeded so far.

Table 2. Proposed radioisotope experiments on time-integrated solar neutrino flux.Tracer Half-life (yr) Target Threshold Reference

(MeV)Ca-41 1.0E5 K-41 236 [47]Kr-81 2.3E-5 Br-81 0.471 [41]Tc-98 4.~E6 Mo-98 1.68 [4s]

pb.~05 1.5E7 T1-205 0.062 [49]

Archaeological and geological dating

4lCa (tln = 100 kyr) tracer could be used to date bones ranging from 50 thousand to1 million years old [50, 51]. This period, while covers an important stage of earlyhuman development, is beyond the reach of lJC-dating. ‘lCa is produced on theground by cosmic rays, therefore, its isotopic abundance varies depending on theexposure length and erosion rate of the ground. Early AMS work indicated that‘lCa/Ca in different bone samples vary in the range of 10-15 to 10-14. However,because the detection limit of AMS, at 2x 10-15,is so close to the measured level thatsystematic errors may contribute to some of the observed variation [52]. If ATTA can

#

push dowm the detection limit to 1x 10“1’, then dwiicated setu s can be used toin~”estigatein detail on the feasibility o f ‘tCa.dating. [n addition. 1 Ca’Ca can be usedto determine exposure age of geological samples.

‘°K?OAr dating is one of the most commonly used of all radiometric techniques[53]. In a variation of this method, the task of measuring “)I@OAr is converted to amuch easier and more reliable one of measuring 39Ar~0Ar. 3qAr is produced by

3qAr(n,p)3qK,and the resulting abundanceirradiating the sample \vith neutrons via thecan be calculated from the known reaction rates and the 39K/”OKratio. Here theisotopic abundance of each argon isotope is quite high (> 10°/0)and mass spectrometryis the standard analysis tool, \vith a detection limit of -107 atoms. ATTA may beusefhl here due to its immunity to contamination [54]. With ATTA, steps of gaspurification could be reduced or eliminated, and thereby minimizing the contaminationby environmental argon. By employing laser excitation to populate argon atoms intothe metastable level, ATTA could lower the detection limit significantly and open upnew applications of 39Ar/ ‘OArdating.

Medical diagnostics

Radioisotope tracers, lJC and 3H in particular, are widely used in biomedicalresearch. With the advancements in AMS, other long-lived isotopes, such as 41Caand2GA1,are also becoming avaiIable to the biomedical field. It is advantageous to uselong-lived isotopes on human subjects due to the low radioactivity of these isotopes.

One particularly interesting proposal that is currently under investigation with AMSis to use 41Ca-tracing to diagnose osteoporosis, a disease commonly found in womenafter menopause [S5]. A patient of osteoporosis loses bones, and hence calcium, at anexcessive rate. The current standard medical practice is to monitor the bone densitywith X-ray imaging. Measuring the bone loss rate directly could be more sensitive tothe patient’s condition, and a faster feedback in assessing treatments. In this proposal,a subject of the high-risk group would take a 4*Ca pill. 41Ca in tissues would bedepleted within a few weeks, but ‘lCa in bones would last a lifetime. 4iCa in urinesamples, with an isotopic abundance in the range of 10-13-10-9,can then be measuredon a regular basis as a monitor of the bone metabolism. Note that AMS has met thetechnical needs of this proposal. The advantage of using ATTA lies in a significantreduction of cost (more than a factor of 10).

Other biomedical applications are nutrition studies using~°K,and 41Ca,and toxic studies using ‘OBeand 2GAI[56].

Mapping ocean currents

Earth climate depends closely on the global ocean currents,

isotopes such as 2%a,

.

which are mapped andmodeled with the help of radioisotope tracers [57]. Radioisotopes are used todetermine the “age” of water, which is the time since the water was near ocean surfaceand exchanged gas with the atmosphere. While the analyses of 3H and lJC have beensuccessfully implemented, 3qAr-tracing is of strong interest because it can be used tomap ocean currents in the range of 10z-103 years and fill the gap between the dating

.ranges of ‘H and 14C. Counting 39Ar atoms in \vater is difficult because 1 liter ofsurface \vater contains only -1x 10~ 39Ar atoms and the isotopic abundance of 39Ar/Aris 8.1x10-16. At present. LLC is used to count ‘9Ar, but r;~uires a kirge amount(-1 000 liters) of \~ater[5S]. AklS has succeeded in detecting Ar at the atmosphericle~’el,but the efficiency needs to be improved [59]. A quick and efficient (-1?40) atom-counting method \vould make a global mapping possible. Since 39Ar/Ar is about 1000times lower than 81Kr/Kr, incremental improvements of our current ATTA system arenot sufficient to realize this application. Ho\vever, populating metastable levels \vithlaser excitation is promising.

Monitoring fission products

Fission products in environment are monitored for a number of reasons. It is usedto assess the contamination of the environment neighboring to nuclear facilities. Itserves as an early warning of nuclear accidents or bomb tests [60]. And it is averification mean to ensure compliance to nucleti nonproliferation treaties,

The fission isotopes that can be analyzed with ATTA include 8%r, 90Sr, ‘33’137CS.

, ACKNOWLEDGMENTS

We thank Carl Wieman and Walter Kutschera for stimulating discussions thathelped initiate this project. We thank the following people for us understand the othertrace analysis techniques: Philippe Collon (AMS), Berrdmrd Lehmann (LLC, RIMS),Michael Paul (AMS), and Klaus Wendt (RHMS). We thank the following people forhelp us understand some potential applications of ATTA: Wick Haxton and KurtWolfsberg (solar neutrino); Paul Renne (39Arf10Ardating); Steward Freeman (41Ca-tracing for osteoporosis); Kenny Gross (nuclear wasteRoland Hohman (39Ar-dating in oceanography).

1)

~)

3)4)5)6)7)8)

9)

10)

REFERENCES

monitor); PhiIippe Collon and

J.R.Arnold and W.F. Libby, Age determinations by radiocarbon content: checks \vith samples of known age,Science 110,678 (1949).?vlethods of low-level counting and spectrometry, International Atomic Energy Agency (IAEA. Vienna. 19S1).Nucl. Instr. Meth. 123(1997), Proceedings of the Seventh [ntemational Conference on AMS.C.Y. Chen et. al. Ultrasensitive iso[ope trace analyses \vith a magneto-optical trap, Science 286.1 i39 (1999).H.H. Loosli andH.Oeschger,37Arand81 Kr in the atmosphere, Earth Planet. Sci. Lett. 7,67 (1969).P. Collon et. al., Measurement of ‘lKr in the atmosphere, Nucl. Instr. and Meth. B 123, 122 ( 1997).H. Oeschger, Accelerator mass spectrometry and ice core research, Nucl. lnstr. and Meth. B29, 196(1987).Isotope of ;Voble Gases as Tracers in Environmental .$hfies, International Atomic Energy Agency (IAEA,1992).F. von l-tippet, D.H. Albright, B.G. Levi. Stopping the production of tissi!e materials for weapons. Sci. Am.253,40 (September, 1985).V.V. Kuzminov ~d A.A. pomansky, New measurement of the 8\ Kr atmospheric abundance, Radiocarbon 22,311 (1980).

.

! i ) K.G. I{cunlann ct. Jl,. Rcixnt dckclopmcnt; in lhcrnml ioni~ation ma~~ sFctlr~m~tri~ tcchniqws for isotopean.rl}sis. Anal)s( 120. 1291 (1995}.

12) R.,-\.\lullcr. Radioisotope dating \\ithac> clotron. Scicncc l96.439(]977J.13) D.E. ?&lson. R.G. Korteling. itnd N’.R. StoU. Carbon-14: dircctdctcction at rmturol concentrations. Science

19s, 507( 1977).[4) C.L. Bcnn~ttct. a\,, Radiocarbon dating using elcctrostztic acw.!erators: ncgati\c ions provide thtkey. Scicncc

, 19S 508( [977).15) P. Collon, Ph.D. Thcsis(Univcrsity of Vienn&Austria l999).16) V.S. Letokhov, Laser ~ho/oioni:ation Spec/roscop.v (/academic press, OrImdo, 1987).17) G.S. Hurst, LI.G. paync. Principles and.4pp/icatiom of Resonance Ionisafion Specfroscop,v(Adm Hilger.

1988).18) G.S.}iurst. &[.G.pa}ne. R.C. Phillips, J.\V.T. Dabbs. B.E. Lehmann. J. Appl. Phys.55. l278(l9S4).19) N. Thonn~d et. a]., Resonmce ionizationspectroscopyandthedetection of81 Kr, Nucl. inst. &leth.B29. 398

(19s7).20) K.zirnmeret.al.. D~terminationof90Srin environmentalsamples with resonanceionizationspectroscopy in

collineargeometry,Appl. Phys. 1359,117(1994).21) B.A. Bushaw and B.D. Cannon, Diode Iaser based resonance ionization masspectrometric measurement of

strontium-90, Spectr. Acts. B52. 1839(1997).22) K. Wendt et. al., Recent developments in and applications of mOrtanCe iOIIiZiMiOTt m= Spectrometty,

Fresenius J. Anal. Chem. 364.471 (1999).23) \~.&f. Fairbmk Jr., photon burst m~ssp~c\rometV+ Nuci. InStr. N[eti. B29,4O7(l987).24) W.M. Fairbarrk et. aL, Photon burst mass spectrometry for the measurement of 85Kr at ambient levels, SPIE

3270, 174 (1998).25) B.D. Cannon ~d T.J. Whitaker, A new laser concept for isotonically selective analysis of noble gase~ AppI.

Phys. B38, 57 ( 1985).26) G. R. Janik et. a!., [sotopically se[ective optical deflection of a krypton atomic beam, J. Opt. Sot. Am. EM,

1617(1989).27) Fujio Shimi.zu, Kazuko Shimizu, Hiroshi Takum~ Chem. Phys. 145,327 ( 1990).28) M. WalhouL H.J.L. Megens, A. Witte, S.L. Ro[ston, Phys. Rev. A48, R879 ( 1993).29) W. Phillips and H. Metcalf, Phys. Rev. Lett. 4S, 596 ( 1982).30) T.E. Thomas, S.W. Dapore-Schw~ M.D. Ray, G.P. Lafayatis, ibid. 67,3483 (199 l).31) E.L. Raab, M. Prentiss, A.E. Cable,S. Chu, D.E. Pritchard, ibid. 59,2631 (1987).32) Z. Hu and H,J. Kimble, Opt. Lett. 19, 1888(1994).33) D. Haubrich et. aL, Europhys. Lett. 34,663 (1996).3-I) B.D. Cannon, Hyperflnd spectra of the radioactive isotopes 81Kr and 85Kr, Phys. Rev. A-47,1148 (1993).35) B.E. Lehman, D.F. Rauber, N. Thonnard, and R.D. Willis. An isotope separator for small noble gas sarnp[e~

Nuci. Instr. and Meth. B28, 571 ( 1987).36) D.J. Vieira et. ai., Trapping radioactive atoms for basic and applied research, to be published in Hyper. Int.37) T. Lee et. al., First detection of fallout G- 135 and potential applications of 137Cs/135Cs ratios, Gee. Cosmo.

Acts 57,3493 ( 1993).38) R. Davis Jr., A.K. Mann, L, Wolfenstein, Solar neutrinos, Annu. Rev. NUCLPart. Sci. 39, 467( 1989).39) J.N. Bahcall and R.K. Ulrich. Solar models, neutrino experiments, and helioseismology, Rev. Mod. Phys. 60,

297 ( 1988).40) W.C. Ha.xton, The SOIWneutrino problem. Annu. Rev. Astrophys. 33,459 (1995).41) G.S. Hurst et. al., Counting the atoms, I?hys. Today 33,24 (September, 1980).42) W.A, Fowler, What cooks with solar neutrinos? Nature 23S, 24 ( 1972).43) A. Cumrning and J.V.C. H&Kton, 3Hc t~spo~ in the sun and the solar neutt-ino problem, Phys. Rev. Lett. 77,

4286 ( 1996).44) D.J. Rokop et. al,, WMS spectrometry of technetium at the subpicogram level. Anai.Qhem. 62, 1271 ( 1990).45) W. Henning et. al., in SOIX Neutrino and Neutrino ~stronomy, 1985, .41P No. 126. edited by hLL. Chew et.

al. (AIP, New York)46) B. Trautmann, Ultratrace analysis of technetium, Radiochimica Acts 63,37.47) W.C. Ha-won and G.A. Cowan, Solar neutrino production of long-lived isotopes and secular variations in the

sun, Science 210, 897 (1980).48) G.A. Cowan and W.C. H~xton, Solar neutrino production of technetium-97 and technetium-98, Science 216.

51 ( 1982).49) MS. Freedman et. al., solar Neutrino: Proposal for a New Test, Science 193, 1117 (1976).50) W. Henning et. al., Calcium-4 I concentration in terrestrial materials: prospects for dating of Pleistocene

sarnp[es, Science, 236.725 ( 1987).51) R.E. Taylor. Dating techniques in archaeology and paleoanthropology, Anal. Chem. 59,3 17A ( 1987).

.

52)53)54)55)

56)

57)58)59)60)

D. Fink. J. Klein xtd R. Middlcton. 4 lCa: past. prcscrtt and futurt. Nucl. Insl. S[cth. B52. 572(1990).Editedby R.&. Taylor and M.J, t\itkcn,Chrorromwic dating in archaeology (Pknum press. New York. 1997).PiIul Rw-me.Berkeley Geochrortology Cerrwr, Berkeley. CA. private communication.D. Elmore m. al.. Calcium-41 as a Iong-twm biological tracw for bone resorption.XUCI. [ns~- kleth. Ml 531( 1990),J.S. Vogel. J. McAninch. S.P.H.T. Frwman. Elements in biological AYLS. NUCL hrstr. hleth. B 12/. 24 I(1997).W.S. Broecker, Chaotic climate. Sci. Am. 62 (November, 1995).H.H. LoosIi,A dating methodwith39Ar, Ear. Plan. Sci. Lett. 63,51 ( [983).W. Kutschera et. al., Long-lived noble gas radionuc[ides, NucI. htst. Meth. B92, 241 (1994).C.R. Carrigan et. a]., Trace gas emissions on geological faults as indicators of underground nuclear testing.Nature 382,528 (1996).

Figure 1. Krypton energy diagrams. (a) Excitation scheme used in the resonance ionizationspectroscopy of k~pton; (b) Excitation scheme used in the laser trapping of krypton;(c) Populatingthemetastablelevel via a non-resonantUV+UVexcitation;(d) Populatingthe metastablelevel via aresonantVUV+lRexcitation.

Figure 2. Signal of a single trapped *lKr atom. The photon counter is only open during the detectionphase. Single atom signal = 1600 photon counts, backgroutld = 340 photon counts.

Figure 3. (a) Fluorescence of trapped krypton atoms. Dark bands are the signal of stable isotopesmeasured with a low-gain photo-diode detector. Line markers mark the positions of the hvo rareisotopes. (b) Fluorescence of trapped %.r atoms versus laser frequency. (c) Number of ‘lFJ and %atoms counted versus laser frequency. Each data point represents the number of ‘ifi atoms counted in3 hours, and ‘5Kr atoms counted in 0.5 hours.

Figure 4. Long-1ived radioisotopes. Isotopes that can be analyzed using ATTA are marked with x.

.

*

Sss.lnm4Ps5s[l/2Jl

I 116.5run

4p6 1

(a)

4PS5P [5/2]2

‘-...,...,.,.. .... .

216.7nm 4ps5s [3/2]*2

............ ............................................. .

216.7nm

(c)

,

4p55p[5’2]’~

-+

8 11.5nrn4ps5s[312]2

.

ElectronImpact

(b)

4Ps5p[3/2][

--t’-........

4P55S[3/2]01 830.0~ ““””’”--.........%..,...

4p’5s [3/2]”2

123.6nm

4p6

(d)

r

25

20

15

10

5

0

I I I I I

One Atom81

t Kr/-.

II1.

f

I

I“Background I

+ \... ....&.,.“““““-”“““““’““?-”;””““““-““........ ....\...............p,. ..................2”{[.............[ (

J /I ~-Le*_-*-*~..,we-_j‘t.=.-b..j.j’‘*.-...,+--@-w %v.”.+-~..”e .. *-<

0.0 ().5 1.0 1.5 2.0

Time (see)

2.5

0.1u7sKr

0.0-1000

1.0

0.8

0.6

0.4

0.2

0.0

353025201510

50

r # I 1 1

86V. I

u8jKr

‘:”r’‘ F(:8’Kr

100 - -400 -20 0 200

~\\

0“’’’’’’’’’’’’’”4

8]Kr (3 hrs.)

85Kr(0.5 hrs.)

.

J

~iJ:L, ....., .... ,. ,.1I * I , I , I , *-150 -120 -90 -60 -30 0 30 60 90

Frequency Offset (MHz)

100 ~

;95 ~ ‘:”nn A

185 ~-

80 ‘ ‘A

AA

A

AAAA

A

A# A AA

A ‘Afi;

A

7570

65n6L ,1

A

&AAA

AAAA

AA

A

AA A#AA

55 q A ‘A

AA

A

A A

A

‘A A

AA

A

A

A

A

A A

&

AA

AA

A AA

—f-h

A

A

AA

A.

.,

Has been -- green arrow ‘ 9can be -- red arrow

AA

AA

A

AAA A AAA

A

AA

‘A

50- AA A

45- ,M AA

40-

35-30-25-20-15-

lo-

5-0 A

1 I I I s 1 1 I I I u

10°I I # I

102I 8

104I I 9 v

106I 1

1081

10’0 10

AA

A

A

A

A“o

A

—DC

AA

Half-Life(year)