Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Science and Technology 570

_____________________________ _____________________________

Accelerator Mass Spectrometry of 129I and Its Applications in

Natural Water Systems

BY

NADIA BURAGLIO

ACTA UNIVERSITATIS UPSALIENSISUPPSALA 2000

Dissertation for the Degree of Doctor of Philosophy in Ion Physics presented atUppsala University in 2000

ABSTRACT

Buraglio, N. 2000. Accelerator Mass Spectrometry of 129I and Its Applications inNatural Water Systems. Acta Universitatis Upsaliensis. Comprehensive Summaries ofUppsala Dissertations from the Faculty of Science and Technology 570. 53 pp.Uppsala. ISBN 91-554-4819-4.

During recent decades, huge amount of radioactive waste has been dumped into theearth's surface environments. 129I (T1/2 = 15.6 My) is one of the radioactive productsthat has been produced through a variety of processes, including atomic weapontesting, reprocessing of nuclear fuel and nuclear accidents. This thesis describesdevelopment of the Accelerator Mass Spectrometry (AMS) ultra-sensitive atomcounting technique at Uppsala Tandem Laboratory to measure 129I and discussesinvestigations of its distribution in the hydrosphere (marine and fresh water) andprecipitation. The AMS technique provides a method for measuring long-livedradioactive isotopes in small samples, relative to other conventional techniques, andthus opens a new line of research. The optimization of the AMS system at Uppsalaincluded testing a time of flight detector, evaluation of the most appropriate charge-state, reduction of molecular interference and imporvement of the detection limit.Furthermore, development of a chemical procedure for separation of iodine fromnatural water samples has been accomplished.

The second part of the thesis reports investigations of 129I in natural waters andindicates that high concentrations of 129I (3-4 orders of magnitude higher than in thepre-nuclear era) are found in most of the considered natural waters. Inventorycalculations and results of measurements suggest that the major sources of radioactiveiodine are the two main European nuclear reprocessing facilities at Sellafield (U.K.)and La Hague (France). This information provides estimates of the transit time andvertical mixing of water masses in the central Arctic Ocean. Results fromprecipitation, lakes and runoff are used to elucidate mechanisms of transport of 129Ifrom the point sources and its pathways in the hydrological environment. This studyalso shows the need for continuous monitoring of the 129I level in the hydrosphere andof its future variability.

Nadia Buraglio, Division of Ion Physics, The Ångström Laboratory, UppsalaUniversity, Box 534, SE-751 21 Uppsala, Sweden

© Nadia Buraglio 2000ISSN 1104-232XISBN 91-554-4819-4Printed in Sweden by University Printers, Uppsala 2000

In addition to the present summary, the Ph.D. thesis consists of the collection of thefollowing papers, which are referred to in the text by the Roman numerals I-V:

I Analytical techniques and applications of 129I in natural water.N. Buraglio, A. Aldahan, G. Possnert.Nuclear Instruments and Methods in Physics Research B, in press, 2000.

II 129I measurements at the Uppsala tandem accelerator.N. Buraglio, A. Aldahan, G. Possnert.Nuclear Instruments and Methods in Physics Research B 161-163, 240-244,2000.

III Distribution and inventory of 129I in the central Arctic Ocean.N. Buraglio, A. Aldahan, G. Possnert.Geophysical Research Letters, Vol. 26, no. 8, 1011-1014, 1999.

IV 129I from the nuclear reprocessing facilities traced in precipitationand runoff in northern Europe.N. Buraglio, A. Aldahan, G. Possnert and I. Vintersved.Submitted to Journal of Environmental Science and Technology.

V 129I in lakes of the Chernobyl fallout region and its environmentalimplications.N. Buraglio, A. Aldahan, G. Possnert.Journal of Applied Radiation and Isotopes, in press, 2001.

© Elsevier Science B.V.: I and II© American Geophysical Union: III

I was involved in the major part of planning and carrying out the experimental work,in the interpretation of the results and in the writing of the papers. I took part in theevaluation and development of the chemical extraction technique and I prepared mostof the sample targets. I performed the AMS measurements at the accelerator facility,as well as data reduction and final calculations.

Conference presentations related to the work presented in the thesis:

European Union of Geoscience 10, Strasbourg-France, 28 March - 1 April 1999.Distribution and inventory of 129I in the central Arctic Ocean, N. Buraglio, A. Aldahanand G. Possnert.

The Second International Congress of Limnogeology, 25-28 May 1999. 129I inprecipitation and runoff in central Sweden and northern Italy, N. Buraglio, A.Aldahan and G. Possnert.

The Second International Congress of Limnogeology, 25-28 May 1999. 129I in freshwater lakes in the area of Chernobyl fallout, N. Buraglio, A. Aldahan and G. Possnert.

Ion Beam Analysis and Accelerators in Applied Research and Technology, Dresden,Germany, 26-30 July 1999. AMS of 129I at Uppsala Tandem Laboratory, N. Buraglio,A. Aldahan and G. Possnert.

Accelerator Mass Spectrometry 8, Vienna, Austria, 6-10 September 1999. Analyticaltechniques and application of 129I in natural water, N. Buraglio, A. Aldahan and G.Possnert.

Abbreviations

In this thesis, the following abbreviations have been used:

ADC Analog to Digital ConverterA.M. Analyzing MagnetAMS Accelerator Mass Spectrometrya.m.u. Atomic mass unitAWL Atlantic Warm Water LayerBNFL British Nuclear Fuel LaboratoryCFD Constant Fraction DiscriminatorDOL Deep Ocean LayerGT1 Groupe Radioecologie Nord-CotentinIAEA International Atomic Energy AgencyL.E. Low EnergyLLNL Lawrence Livermore National LaboratoryMCP Microchannel PlateNAA Neutron Activation AnalysisNIST National Institute for Standard and TechnologyPML Polar Mixed LayerSRM Standard Reference MaterialTAC Time to Amplitude ConverterTL Transition Layer

Contents

AbstractList of papersAbbreviations

1. Introduction 1

2. Sources and occurrence of 127I and 129I 32.1. Stable iodine in nature 32.2. Sources of 129I 52.3. Anthropogenic sources 6

2.3.1. Nuclear weapon tests 62.3.2. The Chernobyl accident 62.3.3. Nuclear reprocessing facilities 7

2.4. The global cycle of 129I 8

3. Chemical analytical procedure 113.1. Extraction of iodine from water 113.2. 129I interlaboratory comparison 133.3. Standards and blanks 143.4. Storage and acidification 14

4. AMS system and measurements 164.1. General 164.2. The ion source 174.3. The injector 184.4. The tandem accelerator 194.5. The post-accelerator system 214.6. The detector system 234.7. Measuring procedure 264.8. Quality assurance 27

5. 129I in marine and fresh water 295.1. Marine water 295.2. Precipitation 335.3. Lakes, rivers and underground water 36

6. Conclusions 38

Appendix. Data analysis 40A.1. Measurement of the ratio of a sample 40A.2. Normalization to the standard 41A.3. Calculation of the numbers of 129I atoms in the sample 42A.4. The background correction 42

References 45Acknowledgements 51

1

INTRODUCTION

Testing of nuclear weapons, nuclear reactor activity and accidents, andreprocessing of nuclear fuel have been among the main sources of environmentalradioactive contamination. These contaminants may be divided into high and lowradioactivity sources depending on their half-lives. The high radioactivity sourcesinclude, for example, 131I (8 days), 3H (12.3 years), 137Cs (30 years) and 90Sr (29years) and represent strong potential health hazards. The low radioactivity sourcesinclude, among others, 14C (5730 years), 36Cl (3.1×105 years) and 129I (15.9 millionyears). Both the former and the latter ones have been used as environmentaltracers. For example, they have been employed for labeling bomb peaks in avariety of archives (sediment and ice), for dating groundwater, and for tracing flowpath movements of water and air masses. In addition to their low radioactivitylevel, the long-lived radionuclides offer the advantage that they can be used toreconstruct the magnitude of radioactivity induced by the short-lived radionuclidesin cases where they have decayed.

129I is a very low source of radioactivity, and thus can be handled as a stableelement that can be used for prediction of pathways of the hazardous 131I.Furthermore, the biophyllic nature of iodine has increased the interest inunderstanding the distribution of the anthropogenically introduced 129I in theearth’s surface environments. Presently, the main sources of this isotope are thedischarges from the two European nuclear reprocessing facilities at Sellafield(England) and La Hague (France) which account for about 95% of the total 129Ianthropogenic inventory (BNFL, 1998; GT1, 1999; UNSCEAR, 1988; Yiou et al.,1995; Raisbeck & Yiou, 1999). This quantity has overwhelmed the natural 129I/127Iratio by 4-5 orders of magnitude.

In spite of the strong interest in studying 129I, a limit was set by theanalytical methods used before the introduction of AMS during the early eightiesas a measurement technique for 129I (e.g. Elmore et al., 1980). Until then, 129I wasmeasured by using the conventional decay counting technique and the NeutronActivation Analysis (NAA) (e.g. Studier et al., 1962). Decay counting requireslarge amounts of sample, in the orders of 105 l compared with the few hundreds ofmilliliters required by the AMS for routine measurements. In addition to being a

2

complex analytical procedure, NAA also requires handling relatively largeamounts of radioactivity.

With the extensive advantage of AMS, a project for measuring 129I wasinitiated at the Uppsala tandem accelerator in 1995 and the first results wereobtained in 1997 (König, 1997). A dedicated beam line, a time of flight detectorand a spherical electrostatic deflector were installed for this purpose. At the sametime, a standard procedure for extracting iodine from water samples was developed(paper I).

With the above described technical arrangements (papers I and II),investigations of the 129I distribution in a variety of reservoirs became possible andthe results are presented in this thesis. The investigations were focused mainly onunderstanding and evaluating the level of 129I in the hydrosphere covering somemarine and continental reservoirs. Most of the radioactive release from thereprocessing facilities enters the North Sea from Sellafield, and the EnglishChannel from La Hague. Swept away by the marine current flow, the waste istransported along the Norwegian coast to the Arctic Ocean. With the aim offollowing and quantifying marine discharges from the facilities through watermass movement, a study of the distribution of 129I in the Arctic Ocean (sampled in1996) was carried out (paper III). Seas and oceans are the major sources of iodinein the atmosphere, in addition to the direct gaseous releases from the facilities.Therefore, monitoring 129I in precipitation can be used to investigate mechanismsof spreading iodine in the environment. Levels of 129I in rain and snow in centralSweden (and in some rain samples of northern Italy) have been recorded for thispurpose since 1998 (paper IV). Further investigations on how 129I is spread intothe hydrological reservoir and on the importance of the different sources (mainlyreprocessing facilities and the Chernobyl accident) were carried out by studyingthe level of 129I in lakes (paper V), rivers and run-off waters (paper I, paper IV)located in central Sweden.

3

SOURCES AND OCCURRENCE OF 127I AND 129I

______________________________________________________

In this chapter, sources and occurrence of anthropogenic 129I in the earth’s surfaceenvironment are discussed. Before introducing these topics, a short review ofstable iodine in nature is presented.

2.1 Stable iodine in nature

The main reservoirs of iodine in the earth’s surface environment are oceans andseas with concentrations in the range 40-65 µg/l. The major forms of dissolvediodine in the marine environment are iodide (I -) and iodate (IO3

-) (e.g. Wong,1991; Hou et al., 1999). Even if the oxidation potential of the iodine-iodate systemin marine waters would favour mainly the presence of iodate, this situation is notalways encountered in reality. It has been shown that in the sea surface wateriodide speciation occurs in significant proportions (up to 50 % of the total iodine)and its concentration decreases with depth (e.g. Takaianagi & Wong, 1986; Wong,1991; Hou et al., 1999). To explain this behaviour, it has been suggested that IO3

–

is reduced to I – at the surface by marine bacteria reducing nitrate (Tsunogai &Sase, 1969) and by diatoms (Wong, 1991). In deeper water, iodate becomes moreabundant (from 50% to 100%). In marine water characterised by strong anoxicconditions, however, such as the Baltic Sea and the Black Sea, iodide is the majorform while iodate is almost undetectable (Wong, 1991; Hou et al., 1999).

From oceans and seas, iodine seems to enter the atmosphere mainly as CH3I(Whitehead, 1984). This assumption is sustained by measurements on ocean watersamples found to be saturated with CH3I that is produced at significant levels inalgae and phytoplankton and abundantly released at the time of mortality(Chameides & Davis, 1980). Another suggested mechanism for iodine to enter theatmosphere is by transformation of iodide into elemental iodine throughphotochemical oxidation (Miyake & Tsunogai, 1963) or through reaction withatmospheric ozone (Whitehead, 1984).

The photochemistry of iodine in the troposphere is not yet fully understood.However, once transferred into the atmosphere, photolytic dissociation of I2 and

4

CH3I most likely produces I atoms through the reactions (Chameides & Davis,1980)

IIhI2 +→ν+ (2.1)and

ICHhICH 33 +→ν+ (2.2)

Possible chemical interactions of I atoms with O3, NOx and HxOx such as

23 OIOOI +→+ (2.3)

lead to the production of several inorganic compounds, that recycle back to I, as,for example,

2NOINOIO +→+ (2.4)

and result in the atmospheric iodine being distributed in gaseous and particulateforms (Chameides & Davis, 1980). The residence time of iodine in the atmosphereis 14, 10 and 18 days for particulate, inorganic and organic gas (mainly CH3I),respectively (Rahn et al., 1976; Whitehead, 1984). Concentrations of iodine in theatmosphere range from a few ng/m3 up to a few hundreds of ng/m3 and average at20 ng/m3 for the global atmosphere (Whitehead, 1984 and references therein).

Re-distribution of iodine from the atmosphere to the earth’s surface (oceansand continents) occurs as dry and wet deposition, with the former taking theleading role in providing iodine to arid regions (Fuge & Johnson, 1986 andreferences therein). Water-soluble iodine species (elemental and particulateiodine, and highly polar HI or HOI) are incorporated into rain and snow oradsorbed to aerosols and removed by precipitation (Chameides & Davis, 1980;Whithead, 1984). In rainwater the concentration of total iodine is usually in therange 1-5 µg/l and the value decreases with the distance from the ocean(Whitehead, 1984 and references therein; Moran et al., 1999). Data concerning thespeciation of iodine in rain still need to be implemented. About 45% of the iodinein some rain collected in the United Kingdom was found to be iodide and 55%iodate (Jones, 1981). However, the concentration of iodate seems to be dependenton the distance from the sea. For example, in a study by Luten at al. (1978) about45% of the total iodine in rain over the Netherlands was present as iodate, but onlyabout 5% for rainwater collected at least 70 km inland from the coast. There isalso evidence that rain might instead contain high concentrations of organic iodine,mainly derived by disintegration of algae and plankton in sea spray (Dean, 1963).

The iodine content in igneous rocks shows a rather uniform concentrationaveraging at 0.24 mg/kg, while sedimentary rocks (i.e. recent sediments,

5

sandstones, shales and carbonates) have a higher concentration ranging from a fewmg/kg up to 200 mg/kg (Fuge & Johnson, 1986). Soils associated with organicmatter, clay, aluminium and iron oxides are usually rich in iodine (mean value = 5mg/kg). Vegetation seems to take up iodine more from the atmosphere than fromsoils and fixes even more iodine upon mortality (Fuge & Johnson, 1986). The pHof the soil influences the form of iodine, as has been shown in several studies (e.g.Fuge & Johnson, 1986; Fuge, 1996). Generally, in acidic soils I - is converted intoI2 which easily evaporates, while in alkaline soils iodine is present as iodates thatdo not escape through volatilisation. Sources of iodine in soils are atmosphericprecipitation, evaporation of subsurface water (especially in arid regions),weathering of bedrock and to some extent agricultural practices (Fuge & Johnson,1986). Losses of iodine from soils are due mainly to removal by groundwater andtransfer to the atmosphere as gaseous iodine.

Iodine concentration in groundwater (mainly present as iodide) is usuallyless than 5 µg/l, with peaks of 50-100 µg/l reported for groundwater with highsalinity or with a high admixture of seawater (Whitehead, 1984 and referencestherein). The concentration of iodine in subsurface brines and thermal and mineralwaters is much higher (up to more than 100 times) than in surface waters, probablybecause of breakdown of organic materials (especially in brines where formationof petroleum takes place) and leaching from sediments (Fuge & Johnson, 1986 andreferences therein).

The concentration of iodine in surface fresh water (rivers and lakes) islower than in seawater. Results of several studies on river water show that thetotal iodine ranges between 2.2 and 42.4 µg/l and the main speciation of iodine isiodide (Wong, 1991 and reference therein; Rao & Fehn, 1999). The concentrationof iodine is also connected with salinity as iodate and total iodine concentrationdrops almost to undetectable level when the salinity is approaching zero (Wong,1991). Studies on lakes have not been so numerous but in some British lakes, bothiodide and iodate have been detected and it seems that there is a conversion ofiodate to iodide during summer (Jones et al., 1984).

2.2. Sources of 129I

Although all the 129I produced during the primordial nucleosynthesis has nowdecayed into 129Xe, 129I is continuously produced by natural processes in theearth’s atmosphere and crust. In the stratosphere, cosmogenic 129I is produced bycosmic-ray-induced spallation of xenon and, in minor quantities, by neutronbombardment of tellurium. In the earth’s crust, the main source of 129I isspontaneous fission of 238U (cumulative yield = 0.027 %). Thermal neutron

6

induced fission of 235U (cumulative yield = 0.54 %) represents another minornatural source of 129I in the lithosphere. Most of the 129I resulting from the fissionprocess arises from the decay of its short-lived precursors, whose half-life rangesfrom 0.99 seconds to 33.4 days.

The amount of natural 129I in the hydrosphere is about 210 kg (1027 atoms),with almost equal contributions from fissiogenic and cosmogenic processes(Fabrika-Martin, 1984). The 129I/127I pre-nuclear-era (before 1945) value in themarine environment has been estimated to be about 10-12 (Fehn et al., 1986).

2.3. Anthropogenic sources

The three main anthropogenic sources of 129I have been bomb tests, nuclearaccidents and reprocessing of nuclear fuel. The release of 129I from these sourceshas increased the natural 129I/127I ratio from 10-12 to as much as 10-5 (Rao & Fehn,1999).

2.3.1. Nuclear weapon tests

Fission of uranium and/or plutonium isotopes in a nuclear explosion leads to theproduction of 129I. It has been estimated that ~ 64 kg (3×1026 atoms) of 129I haveentered the atmosphere as a consequence of surface and atmospheric detonation inthe Northern Hemisphere since the fifties (Rao & Fehn, 1999; Wagner et al.,1996). After 1973, detonations were mainly subsurface and did not contributesignificantly to the 129I dose. The level of the 129I/127I ratio has increased to about10-10 in the Atlantic Ocean as a consequence of the fallout from nuclear weapontests (Edmonds et al., 1998).

2.3.2. The Chernobyl accident

Among the nuclear accidents of the 20th century, the release from the nuclearpower plant at Chernobyl (26 April 1986) was the most serious in terms ofradioactivity. Plutonium isotopes, 123Xe, 85Kr, 131I, 129I, 132Te, 134Cs, 137Cs, 95Zr,103Ru, 106Ru and 90Sr were among the released radionuclides. After the explosion,the most volatile elements such as cesium, iodine and tellurium easily entered theatmosphere. During the days following the accident, the radioactive cloud wascarried by winds towards Scandinavia, contaminating mainly Sweden and Finland,before reaching other parts of Europe. About 1.3 kg (6×1024 atoms) of 129I werereleased into the atmosphere after the Chernobyl accident (Yiou et al., 1994),

7

which is only about 2 % of the total 129I emission from weapon tests. However, therelease was quite well localised in space and time, and some parts of Europe suchas Scandinavia, for example, received considerable doses.

2.3.3. Nuclear reprocessing facilities

Spent nuclear fuel consists of about 95% 238U, a small portion of 235U that has notundergone fission, plutonium and radioactive fission products (among these 129I).In a reprocessing facility, the spent fuel is separated into three components:uranium, plutonium and waste (fission products). The two main nuclear fuel-reprocessing plants in Europe are Sellafield (UK) and La Hague (France). At thepresent, about 99% of the 129I reprocessed is extracted in liquids that aredischarged into the Irish Sea and the English Channel (Baetsle, 1990). The totalamount of 129I liquid discharges is about 789 kg from Sellafield and 1830 kg fromLa Hague up to 1998 (BNFL, 1998; GT1, 1999; Yiou et al., 1995; Raisbeck et al.,1999) giving a total of 2619 kg or 1.2×1028 atoms. Airborne 129I dischargesamount to 57 kg from La Hague until 1996 (GT1, 1999). Nowadays, thereprocessing facilities at Sellafield and La Hague are the main sources ofanthropogenic 129I.

Figure 2.1. Yearly liquid (filled diamonds, from BNFL, 1998 and GT1, 1999, andopen circles from Yiou et al., 1995 and Raisbeck & Yiou, 1999) and atmospheric(open squares) discharges (BNFL, 1998 and GT1, 1999) from nuclearreprocessing facilities at La Hague (a) and Sellafield (b) since 1966.

1

10

100

1965

1970

1975

1980

1985

1990

1995

2000

(b) Sellafield

129 I (

kg)

0.01

0.1

1

10

100

1000

1965

1970

1975

1980

1985

1990

1995

2000

(a) La Hague

129 I (

kg)

8

2.4. The global cycle of 129I

In 1981 Kocher presented a dynamic linear box model of the global 129I cyclebased on available data for stable iodine (to define inventories in the differentsections) and on data on the global hydrological cycle (to define fluxes betweencompartments). In his model the flux (Fi,j) from the ith compartment to the jthcompartment is represented mathematically by the relation

Fi,j = ki,j Yi (2.5)

where Yi is the iodine inventory in the ith compartment and ki,j the fractionaltransfer rate from i to j of iodine per unit time. Therefore, using Fi,j as expressionfor fluxes, the inventory of the ith compartment at a time t can be expressed as

dYi(t)/dt = Ii(t) + j i≠ kji Yj(t) – (

j i≠ ki,j + ) Yi(t) (2.6)

where Ii is the flux of an iodine source (natural or anthropogenic) and is theradioactive decay constant. If, for the description of each nth compartment, we useequation (2.6), the inventory of iodine as a function of time can be monitored ineach section by solving a system of n linear differential equations.

Fabrika-Martin (1984), Liu & Gunten (1988) and Wagner et al. (1996) haveslightly modified the global model of Kocher both by updating inventories andfluxes on the basis of new available data and by dividing some of thecompartments while adding/removing others in order to adapt the model to the aimof their work. Wagner et al., for example, added the stratosphere compartment toaccount for 129I resulting from bomb tests, and removed compartments that are notrelevant to effects occurring on a few-decades time scale.

Considering the model in figure 2.2 and using as input sources theanthropogenic 129I injected by the reprocessing facilities (figure 2.1), the bombtests (Wagner et al., 1996), and the Chernobyl accident (6×1024 atoms of 129I), it ispossible to predict how 129I would spread in the earth’s environment over time. Asstarting conditions, the inventory of 129I (kg) in each compartment has beencalculated by considering the stable iodine inventory and setting the ratio between129I and 127I equal to 10-12 (Fabrika-Martin, 1984; Fehn et al., 1986). Thecalculated values (129I in kg) of each compartment, for the year 1998, are reportedin table 1.1 and the evolution over time in figure 2.3.

It is worth to observe that this model has several limitations. The main oneis the assumption that 129I input would distribute uniformly in the compartmentinto which it is dropped. This is not completely true for the liquid releases fromthe facilities into the ocean. Another limitation is the fact that the residence time

9

and the pre-antropogenic ratio of some compartments may need to be significantlyimproved by measuring natural samples. Furthermore, a local model, with moresuitable transfer fluxes between the boxes, would be more appropriate. Thepresented (figure 2.2) calculation, however, aims at providing a general picture ofhow 129I would be distributed in the different natural compartments and of where itwill most likely accumulate over time.

Stratosphere 2

71 29

Troposphere 49 Troposphere 11 Biosphere(ocean) 0.1 38 (land) 0.09 (land) 20

62 1 51 1 89

Upper Ocean 12 Soil Water 100 Soil 20 0.05 87 1000

99 100

Deep Ocean1000

Figure 2.2. Box model for the global cycle of iodine. The numbers in each boxrepresent the residence time in years. The numbers close to the arrows representthe percentage of iodine leaving each box (Wagner et al., 1996).

Table 1.1. 129I (kg) in each compartment, using the box model for year 1998

OceanAtm.

LandAtm.

Bio-sphere

Soilwater

Soil UpperOcean

DeepOcean

1.12 0.38 0.31 0.11 70.2 1999 877

10

10-5

0.0001

0.001

0.01

0.1

1

10

100

1000

1950 1960 1970 1980 1990 2000

129 I (

kg)

year

ab

defg

h

c

Figure 2.3. 129I inventory in different compartments from the Europeanreprocessing facilities, bomb tests and Chernobyl accident calculated using thebox-model in figure 2.2.

a = upper oceanb = deep oceanc = soild = biospheree = ocean atmospheref = land atmosphereg = soil waterh = stratosphere

11

CHEMICAL ANALYTICAL PROCEDURE

In order to be able to perform AMS measurements of 129I, iodine has be extracted fromnatural samples and converted to silver iodide. In this chapter, extraction of iodinefrom natural water is described and discussed.

3.1. Extraction of iodine from water

In this study, the chemical extraction of iodine from water samples exploits theproperty of molecular iodine, I2, to be soluble in carbon tetrachloride, CCl4, whileiodide, I –, is soluble in water (Moran et al., 1995). The procedure used for extractingiodine is the following (figure 3.1). A water volume of 50-1500 ml is passed through a0.45 µm glass fibre filter and mixed with 0.5-2 mg of iodine carrier, in the form ofpotassium iodide, KI, or I2. The amounts of carrier and of water used depend on thenatural isotopic ratio of the sample, and they are combined in order to obtain anisotopic ratio of one or two orders of magnitude higher than the analytical background(10-14-10-13). Typically, 2 mg of iodine carrier is dissolved in 200 ml of water sample.As pointed out in an earlier chapter, iodine in water is present mainly as iodide andiodate, and therefore it is important to ensure that there is isotopic equilibrium between129I (often both in the iodide and iodate forms) and the carrier (iodide form). In orderto achieve this, 2 ml of 0.1 M sodium sulfite hydrogen, NaHSO3, is added to thesample to convert all the iodate into iodide. The sample is then acidified with nitricacid, HNO3, transferred into a flask where 15 ml of CCl4 and 4 ml of the oxidizingagent water peroxide, H2O2, are added. After shaking, the iodine gets into the organicphase forming a purple solution that is transferred in a clean beaker. This purificationstep is repeated several times until, after the addition of CCl4, the pink color is notobserved, indicating that most part of the iodine has been extracted from the waterphase. The pink organic phase is now transferred into another clean separation funneland reduction of iodine is achieved by addition of 10 ml of 0.1 M NaHSO3 and of afew drops of 50 % sulfuric acid, H2SO4. The solution is then transferred into a cleantube and 1.5 ml of 0.01 M silver nitrate, AgNO3, is added to precipitate silver iodide,AgI. An excess of AgNO3 is not recommended because undesired silver-sulpha

12

compounds precipitate. Washing of AgI with super-pure ammonia NH3 followed by awash with double-distilled water is carried out in order to dissolve any silver chloride,AgCl, or small amounts of the silver-sulpha compounds. At this point, the tubecontaining AgI is covered with aluminium foil and placed in an oven (50º C) to dryovernight. The recovered AgI is mixed with a double amount in weight of niobium orsilver powder and afterwards pressed into a copper holder for AMS measurements.The recovery of AgI is usually better than 70 % and the major losses most likely occurduring the precipitation of AgI and the washing stages.

Water

Addition of carrier

NaHSO3

Transformation of iodatesHNO3 (pH = 2) into iodides H2O2

Oxidation of I- into I2

CCl4

Extraction of I2

NaHSO3 H2SO4 Reduction of I2

into I-

AgNO3

Precipitation of AgI

Figure 3.1. Flow chart showing the basic steps of chemical extraction of iodine fromwater samples.

13

The step performed to ensure isotopic equilibrium (early addition of NaHSO3)was made during a later stage of the project. In the beginning, in fact, a slightlymodified procedure, where iodide and iodate were recovered separately, was used.After recovery of iodide, any possible iodate was transformed into I2 by using 5 ml of1 M hydroxylamine hydrochloride, NH2OH.HCl. Molecular iodine was then dissolvedinto CCl4 and collected. After this step, the procedure continued as the one previouslydescribed. In order to compare the two methods, the same water was preparedaccording to the two approaches. The results (table 3.1) agree within the 5%measurement error (paper I).

Table 3.1. Results of 129I/127I from the same samples prepared using the two chemicalextraction procedures described in the text. A = with addition of NaHSO3 beforeacidification by HNO3, B = with addition of NH2OH.HCl. The snow sample and theriver Fyrisån water were collected in Uppsala (Sweden)

SAMPLE A B

Baltic Sea (40 m) (1.3±0.07)×10-11 (1.2±0.07)×10-11

Baltic Sea (100 m) (5.4±0.3)×10-11 (5.5±0.3)×10-11

Snow (991212) (6.9±0.3)×10-12 (6.8±0.3)×10-12

Fyrisån (000202) (5.1±0.3)×10-12 (5.2±0.3)×10-12

North Sea (2) (6.5±0.2)×10-10 (6.0±0.3)×10-10

North Sea (3) (7.2±0.3)×10-10 (7.7±0.3)×10-10

3.2. 129I interlaboratory comparison

An interlaboratory comparison exercise on 129I was organized and conducted by theLawrence Livermore National Laboratory (LLNL) on behalf of the InternationalAtomic Energy Agency (IAEA). A total of eleven samples were delivered to thelaboratories participating in the exercise. Four samples were synthetic standard AgI(LL samples), one was water containing iodine in the form of KI (250 mg I per gramof water) and the others were environmental materials (pine needles, maple, weeds,soils).

The iodine in the water from the interlaboratory comparison was extractedusing the procedure described above. The result is in good agreement with valuesreported by the other laboratories (table 3.2).

14

Laboratory 129I atoms/gram

Uppsala (1.21±0.09)×107

B (1.20±0.03)×107

W (1.22±0.06)×107

G (1.21±0.08)×107

K (1.83±0.05)×107

M (1.12±0.05)×107

3.3. Standard and blanks

AgI powder precipitated from a bulk solution obtained after successive dilutions ofStandard Reference Material (SRM) 4949C.Iodine 129 (National Institute for Standardand Technology, NIST) was used in this study. The standard was prepared withseveral isotopic ratios, namely (0.21±0.02)×10-11, (1.17±0.02)×10-11 and(5.00±0.17)×10-11. In most cases, the standard with isotopic ratio 1.17×10-11, preparedand delivered by the late Dr. Kilius from the Isotrace Laboratory in Canada, was used.The other two standards (prepared at Uppsala) were used for measurements of sampleswith higher or lower ratios than 10-11.

The blanks (129I free samples) used in this study include super-pure KI,Woodward iodine (I2) and a natural Precambrian (> 300 Ma old) iodargyrite crystal(AgI). The former two materials were also used as carrier. The 129I/127I value of theblanks indicates that iodargyrite crystal has the lowest ratio (see appendix).

3.4. Storage and acidification

For logistic reasons, the samples were sometimes stored for weeks or months beforebeing chemically treated. This may introduce changes in the original 129Iconcentration due to adsorption on the walls of the container and on particles, or due tobacterial growth and photolysis reactions. To reduce these effects, the water samplesin polyethylene bottles were stored in a dark and cold (4o C) room. Several tests weremade to evaluate the effect of storage (paper I) and the results are summarized in table3.3. Although the storage effect seems to vary depending on the type of water (fromoceans, lakes, rivers etc.), the results show that leaving the sample without anytreatment and in a dark cold room does not change the 129I concentration more than10 % irrespective of the storage period (up to 15 months). Among the several factorsthat could eventually modify the samples upon storage, we checked the effect of pH,with an experiment carried out on river water (river Fyris in Uppsala) having a naturalpH of about 7. A sample of the river water was portioned to 8 equal volumes and the

Table 3.2. Results of 129I in thewater sample used in theinterlaboratory comparison exercise.The letters refer to the differentlaboratories that participated in theexercise (Roberts et al., 1997)

15

pH was adjusted to values between 3.5 and 8.5 through the use of HNO3 and sodiumhydroxide, NaOH. These portions were kept at constant pH in closed beakers at roomtemperature for 12 days and then analyzed. The results of this test indicate that forpH < 7 the 129I concentration increases as the pH decreases and this most likely is dueto release of iodine by dissolution of particulate matter (paper I; Kaplan et al., 2000).Further results on the effects of acidification and filtration are reported in table 3.3.

Table 3.3. 129I (atoms/liter) in samples stored at different conditions. Water from theFyris River was stored after filtration and acidification (pH = 3) for 3 months. FyrX(where X is a number) stands for water collected at different locations along thecourse of the river. Fyr6 has been stored for 10 months after acidification. Lakes 1and 2 have been stored for 15 and 11 months, respectively, without any treatment. TheBaltic Sea samples have been stored for 1 month without any treatment (paper I)

After sampling After storage

Fyr1 (2.2 ±0.2)×108 (2.8 ±0.3)×108

Fyr2 (2.0 ±0.2)×108 (2.3 ±0.3)×108

Fyr3 (2.8 ±0.3)×108 (2.8 ±0.4)×108

Fyr4 (2.9 ±0.2)×108 (3.2 ±0.4)×108

Fyr5 (3.4 ±0.3)×108 (3.8 ±0.3)×108

Fyr6 (3.8 ±0.3)×108 (4.6 ±0.3)×108

Lake1 (3.3 ±0.2)×108 (3.3 ±0.1)×108

Lake2 (4.9 ±0.3)×108 (5.1 ±0.2)×108

Baltic1 (3.0 ±0.1)×109 (3.1 ±0.1)×109

Baltic2 (3.8 ±0.1)×109 (4.0 ±0.1)×109

Baltic3 (2.6 ±0.1)×109 (2.9 ±0.1)×109

16

AMS SYSTEM AND MEASUREMENTS

In this chapter, the AMS technique for 129I and the Uppsala tandem accelerator facilityare described.

4.1 General

The 129I AMS measurements were performed at the Uppsala Tandem Laboratory. Theaccelerator system (figure 4.1) is a shared facility where 14C, 10Be and 129I are routinelymeasured for 50% of the time while the remaining time is dedicated to ion beammaterials analysis and modification besides basic studies of ion interactions withmatter.

Figure 4.1. Drawing of the Uppsala tandem facility, where A is the accelerator hall,hosting the accelerator high-pressure vessel (yellow), B is the control room and C isthe experimental hall. For an approximate scale, the size of the yellow tank is 11.5 m× 1.5 m.

17

4.2 The ion source

The ion production in a system used for AMS must fulfill special requirements(Middleton, 1984; Kilius et al., 1997a) mainly related to the fact that the isotopic ratiosto be measured can be very low (down to 10-15 or less). With respect to this, cesiumsputter ion sources are the most commonly used in AMS. The main reason for this isthat they have negligible memory effects. A schematic layout of the cesium sputteringion source at the Uppsala Tandem Laboratory is shown in figure 4.2.

Figure 4.2. Principle drawing of the cesium sputtering ion source showing theprinciples of the negative ion beam current extraction ( ström & Possnert, 1983).

Metallic cesium is contained in a reservoir cavity and heating the cavity to atemperature of about 200-300 °C produces cesium vapour. Cesium is diffused througha porous tungsten crystal (kept at a temperature of about 1100 °C) where the cesiumatoms undergo surface ionization. The ionizing electrode is kept at a potential ofabout 6 kV to accelerate the cesium beam towards the sample target, while theaccelerated beam is focussed and steered by two Einzel lenses. The central electrodeof each Einzel lens is divided into four parts allowing focussing and steeringsimultaneously. When the cesium ions collide with the sample surface, sputteringejection of neutral atoms and positive/negative ions takes place. The negative ionbeam current is strongly dependent on the ionization. In the most favorable cases,such as for carbon, iodine and chlorine, ionization efficiency in the range 5-8 % can beobtained. For other elements, such as beryllium and aluminum, not easily formingnegative ions, molecular beams BeO- and AlO- are preferably used. The extracted

Cs reservoir

Ionizer

Einzel lens

Einzel lens

Einzel lens

Einzel lensSteering

Towardslow energydeflector

Sample holder~ -30 kV

~ + 6 kV

18

negative ion current from the cesium sputter ions source is typically in the order of 1-10 µA, allowing a statistical precision of about 1-5 % within 1 hour of measuring time(assuming an accelerator transmission of about 20 % and charge state +1, seefollowing paragraphs).

In order to have stable measurement conditions, a beam emittance smaller thanthe acceptance of the subsequent optical elements (flat top) has to be achieved. Forthis reason, among other things, asymmetric cratering on the target during thesputtering process has to be avoided.

The negative ion beam is extracted from the ion source at an energy of about 30keV and focussed to a waist at the slits before the electrostatic deflector (figure 4.3).

Figure 4.3. Schematic drawing showing pertinent ion optical elements of the lowenergy side of the accelerator system.

4.3. The injector

A schematic drawing of the injector and the low energy side of the accelerator isshown in figure 4.3. The negative ion beam produced in the ion source has a spread inenergy of the order of tens of eV with a tail at the high-energy side. Besides negative129I ions, such molecules as 128TeH- and 127IH2

-, the isobar 129Xe-, 127I- and 128Te- ionsare part of the beam. The number of interfering ions and molecules is several orders

Ion source

Einzel lensesSlits

0.5 m0.4 m0.4 m

0.4 m

0.71 m

0.71 m 0.39 m 0.71 m

Slits

Slits

Pre-accelerationtube (75 kV)

Einzel lens

2.769 m

Electrostaticdeflector

Injectormagnet

Steering

L.E. cup

Injector cage

Acceleratortank

19

of magnitude higher than the number of 129I, making the detection of the element ofinterest impossible, without the use of an accelerator.

In order to suppress the unwanted ions and molecules, an electrostatic andmagnetic separation (the physics of which is based on the Lorentz formula,

Eq)Bv(qF +×= ) is performed. The double focussing 90° spherical electrostaticdeflector (radius = 0.4 m and electrode distance = 40 mm) is dispersive in energyaccording to

)r(qE ε=

21 (4.1)

where E and q are the energy and the charge (-1) of the beam, respectively, ε the valueof the electric field, and r the electrostatic deflector radius. Part of the high-energy127I-, 128Te- ions that would mimic 129I- ions after the injector magnet are thus rejectedby the dispersive action of the electrostatic deflector. The ion beam trajectory in the90° double focussing fringing field dipole injector magnet (bending radius = 0.376 mand maximum value of the magnetic field = 1.1 T) follows the equation

qB

)mE(qBp

qBmvr

/ 212=== (4.2)

where m is the mass, v the velocity, q the charge, p the momentum of the particle andB is the magnetic field strength. The magnetic deflector is thus dispersive with respectto the momentum.

Theoretical ion optic calculations and measurements showed the importance ofthe electrostatic deflector in suppressing most of the interfering ions or molecules(figure 4.4). The suppression of 127I relative to 129I ions by the injector is about 3orders of magnitude (König, 1997).

Before entering the tandem accelerator, the beam is pre-accelerated to 75 kV inorder to reduce the strong lens effect at the entrance of the accelerator. A smallerwaist at the stripper channel and an improved transmission is thereby obtained.

4.4. The tandem accelerator

The tandem Van der Graaf machine is a type of particle accelerator in which the high-voltage at the terminal (nominal value = 6 MV) is used twice to increase the energy ofthe injected ions. Negative ions are accelerated by the positive potential of theterminal. When reaching the terminal, electrons are stripped off in a thin foil or gas,

20

and a second acceleration takes place by repulsion back to ground potential (thetandem principle). Molecular interference is eliminated by Coulomb explosion afterthe charge exchange process in the stripper, if the charge-state of the molecules ishigher than +2. The high energy obtained after the acceleration allows for a uniqueidentification of each ion by conventional nuclear detection techniques such as E-∆Eor time of flight-energy detector systems.

Figure 4.4. Ion current measured after the slits (before the Low Energy (L.E.) cup infigure 4.3) for different masses before and after the installation of the sphericalelectrostatic deflector (König, 1997).

The charge state distribution of a beam of ions passing through a gas or foildepends on the velocity and type of the ions and on the stripper composition (Betz,1972). The charge state distribution for chlorine and iodine, acquired at our system fordifferent values of the ion velocity, is shown in figure 4.5. In order to have the highestpossible count rate at the detector, the charge state to be chosen should be the mostprobable one. However, this choice is limited by other factors such as the magneticrigidity of the analyzing magnet and the possible presence of interfering molecularfragments.

Mass (a.m.u.)

without electrostaticdeflector

with electrostaticdeflector

21

Figure 4.5. Charge state distribution for chlorine (a) and iodine (b) measured atdifferent energies for an argon gas stripper. The current is measured at the AnalyzingMagnet (A.M.) cup and it is normalized to the L.E. cup current.

4.5. The post-accelerator system

At the exit of the accelerator tube, the energy of the ions is given by the equality

)qMm(eVE += (4.3)

where e is the elementary charge, V is the terminal voltage value, m is the mass of theion, M is the injected mass of the ion or molecule and q is the charge state. The beamis deflected through the 90° double-focussing fringing analyzing magnet and deflectedthrough the 30° switching magnet before entering the 20° cylindrical electrostaticdeflector and the detection region (figure 4.6). The bending radius (r = 1.016 m) andthe limited maximum value of the magnetic field (B = 1.6 T) in the analyzing magnetrestrict the charge state that can be selected at a fixed energy of the ions, according toequation 4.2. For iodine, for example, it is not possible to select a charge state lowerthan +4 because that would require an analyzing magnetic strength that our presentsystem cannot reach.

After acceleration and stripping, the beam, in addition to ions of interest,consists of different components, such as background isobars, molecular ions and/ormolecular fragments. Some of these latter ions have the same magnetic rigidity as

0

10

20

30

40

50

0 1 2 3 4 5 6 7 8 9 10

T.V. =4.5 MVT.V. = 5.0 MVT.V. = 5.2 MVT.V. = 5.5 MV

norm

aliz

ed c

harg

e sta

te d

istri

butio

n (%

)

charge state

(a) chlorine

0

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.083.0 MeV3.5 MeV4.0 MeV4.5 MeV

3 4 5 6 7 8 9 10 11charge state

(b) iodine

rela

tive

char

ge st

ate

distr

ibut

ion

22

129I+5 ions and, therefore, cannot be separated solely by magnetic deflection. Anadditional electrostatic separation is therefore applied (figure 4.6), resulting, for iodine,in a suppression of 127I relative to 129I of about 7 orders of magnitude (König, 1997).According to the diagram of figure 4.7 (Purser et al., 1979), all ions having the samem/q value as the 129I+5 ions will then reach the detector system. In order to uniquelyseparating ions having the same m/q one has to employ an energy detector ( E-E gascounter or solid state), a time of flight-energy detection system and an absorber or agas-filled magnet.

Figure 4.6. Schematic drawing showing pertinent ion optical elements of the high-energy side of the accelerator system.

An example demonstrating the problem of distinguishing ions with the samem/q is the following:

( )

( ) eV271

18eV23O

qEM

eV23)1

3618(eVO

qE

182

18

==

=+=

+

+

(4.4)

If the charge state +2 is chosen for 36Cl then

1.22 m 3.5 m 2 m

2 m

0.59 m

0.4 m

4.21 m

2.02 m

1.39 m

High energyquadrupoles

Switchingquadrupoles

Beamlinequadrupoles

Switchingmagnet

Analyzingmagnet

Electrostaticdeflector

ToFEnergydetector

Slits

Slits

Slits

Acceleratortank

23

( )

( ) eV27236eV

23Cl

qEM

eV23)

221(eVCl

qE

2362

236

==

=+=

+

+

(4.5)

Figure 4.7. m/q plotted against the variable E/q, showing a hyperbolic linerepresenting magnetic selection, a vertical line representing electrostatic selection, anhorizontal line corresponding to m/q = constant and a straight line through the originrepresenting particles of constant velocity.

129I presents a similar problem when charge state +4 is chosen, due to 97Mo+3

derived from molecular dissociation of 97Mo16O2- (Kilius et al., 1997b). It is therefore

preferentiable to employ other charge states. With our system, we have evaluated +5and +7 and have chosen +5 because of the higher 127I current and lower backgroundinterference (figure 4.8).

4.6. The detector system

For 129I the main sources of interference are 128Te and 127I while the isobar 129Xe- doesnot represent a problem because it is not metastable (e.g. Boaretto et al., 1990). Mostof the 128Te- and 127I- ions are likely to be eliminated already at the injector, while128Te- and 127I- ions originating from molecular dissociation in the stripper represent

1+2+3+4+

mE/q2 = const.

analizingmagnet

v = const.

E/q

m/q

m/q = const.

24

the main source of interference at the detector. By using the m/q relation, possibleinterfering ions can be predicted. For charge state +5 the following ions representpotential interference:

q = 1 M = 25.8 a.m.u. (26Mg+1)q = 2 M = 51.6 a.m.u. (52Cr2+)q = 3 M = 77.4 a.m.u. (77Se3+)q = 4 M = 103.2 a.m.u. (103Rh+4)

Figure 4.8. Time-energy spectra recorded by the acquisition system when the selectedcharge state was +5 for a target having an 129I/127I ratio of 10-11.

When silver was used as a binder (see paper II and chapter 3), 103Rh+4 represented anundesirable interference to 129I, while no interference was detected when niobium wasemployed (figure 4.8).

In order to identify the different ions, a combination of time of flight andenergy detection is employed (velocity filter). The basic idea of the time of flight-energy detector system is shown in figure 4.9. The time detectors are made of thincarbon foils (5 µg/cm2 of thickness and 20 mm in diameter) in which secondary

180 190 200 210 220 23050

60

70

80

90

100

110

120

130

140

150

Time

Ene

rgy

I-129

I-127

25

electrons are produced when the ion beam passes through. The created secondaryelectrons are thereafter accelerated by a grid system (figure 4.10) and deflectedtowards the two Microchannel Plates (MCP). The function of the MCP is to amplifythe signal. For each output pulse from the MCP that is above the Constant FractionDiscriminator (CFD) threshold, each CFD creates a fast

Figure 4.9. Principle drawing of the time of flight and energy detector. The time offlight information is produced by the signals delivered by two carbon foils (the startand stop detector) when the ions pass through. The ions reaching the energy detectorare counted as significant events if they are in coincidence with the signal from thetime of flight detector. The time of flight difference between 129I (286.4 nsec) and 127Iis of 4.43 nsec for our present detector setup and for ion beam energy of 21 MeV.

signal (figure 4.11). The time difference between the signals from the start and stopdetectors is converted by the Time to Amplitude Converter (TAC) to an analog pulsethat is analyzed by the ADCs in the acquisition system. A silicon-charged particledetector performs the energy analysis.

The overall efficiency of the detector system is limited by the efficiency of thetwo time of flight detectors, while the energy detector can be assumed to be 100%efficient. The efficiency of one time of flight detector has been measured to be about60% for alpha particles of 5.5 MeV (König, 1997).

L = 1.6 m

Carbon foils detectors Energy detector

Beam

Start Stop

t

ToF telescope

0.22 m

EnergyTime

26

Figure 4.10. Electrostatic-mirror carbon foil time detector (after Busch et al., 1980).The advantage of this arrangement is isochronity, meaning that the time of flight fromthe foil to the channel plates is independent of the point from which the electronsoriginated.

4.7. Measuring procedure

During routine sequential measurements, 129I particles are counted for 300 seconds atthe detector and the stable isotope 127I current is measured for 5 seconds at the Faradaycup after the analyzing magnet. The 129I/127I ratio of the sample is calculated bydividing the 129I counts by the average between the two measured 127I values of thecurrent, in order to compensate possible changes in 127I current delivered from thesample. Each of these measurements (cycles) is repeated at least three times beforeswitching to the next sample. The sequence in which the samples are measuredincludes a standard sample, 4 unknown samples, and a standard sample again. Theobtained 129I/127I ratio of each unknown sample is normalized to the 129I/127I ratiomeasured for the standard whose nominal value is known. The value of the relativeratio is used to calculate the amount of 129I in atoms/liter in the sample. The procedureadopted for the data analysis is presented in the appendix.

C foil

Field freeregion

Anode

MCP1

MCP2

Electrostaticmirror

Coaxial connector

grid

Beam ine-1

e-1

20 M Ω

20 M Ω

15 M Ω

15 M Ω

5 M Ω 50 M Ω

27

Figure 4.11. Electronic set-up of the detector system for the data acquisition.

Typical currents of 127I at the analyzing magnet Faraday cup range at 300-600nA when charge state +5 is selected and the corresponding accelerating transmission is4%. Precision as good as 3% is normally achieved for samples having a ratio of theorder of 10-12-10-11. One or two blanks are regularly measured to check the overallbackground in the system. The system background is limited to 4×10-14 (seeappendix).

4.8. Quality assurance

The performance of the Uppsala AMS system with respect to application of 129Imeasurements was controlled through an inter-comparison test carried out incollaboration with the LLNL and Zurich-Hannover laboratories (see section 3.2). Ourmeasurements of the LLNL exercise are comparable with the reported expected valuesand with the ranges of the results of the other laboratories (table 4.1). Also the resultsof measurements of the samples chemically prepared at Hannover University showgood agreement between the two laboratories (table 4.2).

EToF 1 ToF 2

delay

ConstantFraction

ConstantFraction

TAC

Pre-Amp

Main-Amp

ADC 2

Pre-Amplifier

Pre-Amplifier

ADC 1

HV HV

Stop Start

Delay-Amp

28

Table 4.1. Results of measurements on the AgI samples prepared for the 129I inter-laboratory comparison. The table reports 129I/127I ratios measured at Uppsalacompared with the ranges of the ratios measured by the other laboratories that joinedthe exercise (Roberts et al. 1997). The estimated expected values are based on thedilution of the NIST standard

Uppsala Range of other laboratories

expectedvalue

LL1 (8.26±0.17)× 1110− (8.20-9.40) × 1110− 9.03× 1110−

LL2 (4.20±0.097)× 1110− (4.22-4.80) × 1110− 4.55× 1110−

LL3 (2.09±0.077)× 1110− (2.03-2.95) × 1110− 2.17× 1110−

LL4 (0.46±0.017)× 1110− (0.48-0.53) × 1110− 0.492× 1110−

Table 4.2. 129I/127I ratio in samples measured at the ETH tandem accelerator facility inZurich and at our laboratory. The samples were chemically extracted at the Zentrumfür Strahlenschutz und Radioökologie of Hannover University

Code Zurich Uppsala

SW5Ø12D (1.46±0.19) ×10-11 (1.39±0.04) ×10-11

SW0Ø28A (4.29±0.18) ×10-11 (3.94±0.18) ×10-11

SW5063A (9.72±0.49) ×10-11 (9.50±0.21) ×10-11

29

129I IN MARINE AND FRESH WATER

Part of the present study has been devoted to investigations of 129I in a variety of naturalwaters. These include the central Arctic Ocean (paper III) and Baltic Sea (paper I),precipitation over Sweden and Italy (paper IV) and some surface waters such as lakes(paper V), rivers and creeks (papers I, IV) located in central Sweden and northern Italy.In this chapter, a summary of these investigations is presented.

5.1 Marine water

During the recent decades, the reprocessing facilities at Sellafield and La Hague havebeen continuously discharging liquid effluents in the Irish and English Channel,respectively (BNFL, 1998; GT1, 1999; Yiou et al., 1995; Raisbeck & Yiou, 1999). The129I discharges from Sellafield were rather constant before the nineties and they haveincreased of 3-4 times by 1998 while discharges from La Hague have increasedexponentially since 1966. Numerous measurements of 129I on seaweed and seawater inthe North Atlantic have been performed since 1991 (Yiou et al., 1994; Raisbeck et al.,1995; Raisbeck & Yiou, 1999; paper I; Hou et al, 2000) in order to improve estimates of129I discharges and to evaluate the possibility of using 129I as a time marker and a tracer ofwater masses movement. Reported marine concentrations in the vicinity of Sellafield andLa Hague were of the order of 1011 atoms/liter (or of 10-6 for the 129I /127I ratio) in 1991-1992 (Yiou et al., 1994) but are presumably much higher at the present time. From thetwo point sources, water streams drive the liquid effluents into the North Sea, which isnow already heavily contaminated as shown by recent studies. The same highconcentrations, measured outside the facilities in 1991-1992, were found, indeed, in thewaters of the North Sea in 1999 (paper I; Szidat et al., 2000). Apparently, from the NorthSea, a small portion of water enters the Baltic Sea (Paper I; Hou et al., 2000). Apreliminary sampling survey of the North and Baltic Seas indicates a gradient in the 129Iconcentration in the surface water, with the lowest value at latitude 61.05˚ N andlongitude 19.35˚ in the Baltic, and the highest in the North Sea (figure 5.1).

30

Figure 5.1. Map showing the concentrations of 129I (atoms/liter) measured in 1999 inseveral locations of the North and Baltic Seas. Filled circles represent ourmeasurements, while filled squares have been measured by Hou et al., 2000. The value of4×1011 atoms/liter was measured by Szidat et al., 2000.

From the North Sea the path of the discharges follows mainly the general watermasses movement towards the Barents Sea along the Norwegian Coast (129I /127I of 4×10-8

in Bergen and 1×10-8 in Kirkenes, Raisbeck et al., 1995).The investigations of the Arctic Ocean (figure 5.2) indicate 129I concentrations of

the order of 108-109 atoms/liter that are much higher than the natural background (105-106

atoms/liter) level (e.g. Kilius et al., 1995; Beasley et al., 1995; Smith et al., 1998; paperIII). Besides the discharges from the reprocessing facilities, Russian riverine (Ob,Yenisey and Lena rivers) input (Moran et al., 1995), direct dumping from submarinenuclear reactors (Strand & Holm, 1993) and nuclear bomb tests at Novaya Zemlyarepresent other sources of radioactive waste into the Arctic.

In particular, data from the Arctic Shelves display a decreasing trend in 129Iconcentration in the surface water from about 109 atoms/liter in the Barents and Kara Seas(Raisbeck et al., 1993; Smith et al., 1998; Josefsson, 1998) to about 108 and 107 in theEast Siberian (Josefsson et al., 1995) and Chukchi Sea (Kilius et al., 1995; Beasley et al.,1995), respectively. This seems to agree with the pattern in the Norwegian Sea (figure5.3) that reflects intrusion of Atlantic water masses from the Norwegian Coastal Currentinto the Barents Sea through the Fram Strait (e.g. Smith et al., 1990). Once the watermasses have entered the Arctic Ocean, the circulation proceeds eastward and splits into

31

the Eurasian and Canadian Basins at the Lomonosov Ridge to merge again towards theFram Strait.

Figure 5.3. 129I /127I values in surface water, sampled in 1993, along the Norwegiancoast from Bergen to Kirkenes (Raisbeck et al., 1995) and in the seas of the Arctic Ocean,sampled in 1994 (Josefsson, 1998).

Figure 5.2. Map of the ArcticOcean. The white area showsthe permanent ice cover andblack arrows illustrate icemovement (Hilmer et al.,1998). The gray arrows showthe surface water circulationfrom the Eastern NorthAtlantic and the central Arctic(paper III).

Berg

en

0

1 10-8

2 10-8

3 10-8

4 10-8

5 10-8

129 I/12

7 I

Location

Kir

kene

s

Norwegian coast

Kar

a Se

a

Lapt

ev S

ea

East

Sib

eria

n Se

a

Arctic Ocean

Bare

nts S

ea

32

This circulation pattern has strongly controlled the 129I distribution in the centralArctic and is well illustrated in paper III. During the Arctic Ocean Expedition of August1996, water samples were collected in the central Arctic at three different stations, namely064 (Lomonosov Ridge, latitude 86-25.0, longitude 144-29.6), 069 (Makarov Basin,latitude 86-10.3, longitude 155-00.3) and 081 (Eurasian Basin, latitude 82-29.9, longitude134-36.3) (figure 5.2).

0 2 108 4 108 6 108 8 108 1 109 1.2 10 9

0

500

1000

1500

2000

2500

3000

3500

4000

Lomonosov Ridge (064)Makarov Basin (069)Eurasian Basin (081)

129 I (atoms/liter)

Dep

th (m

)

halocline

DO

L

PML

TL

The results (figure 5.4) indicate that the highest 129I concentrations (> 70×107

atoms /liter) occur within the first 100 m from the surface and that the 129I concentrationdrops gradually and reaches values of 20×107 atoms/liter at depth of 800-1000 m.Below 1300 m, the concentration drops as low as 3×107 atoms/liter while at depths ofabout 3000 m and 4000 m, the 129I values drop to about 7×106 atoms/liter. 129I /127I ratiomeasured for the deepest samples are one order of magnitude higher than the pre-anthropogenic predicted background (table 5.1). This suggests that the anthropogeniccontamination may have already reached the deepest part of the Arctic Ocean.

An attempt is made in paper III to evaluate the transport pathways from the sourcesto the Arctic. The results show that > 95% of the 129I was transported through oceancurrents and the rest through riverine input and direct precipitation. Based on thisconclusion and on a calculation of a total inventory of 2.7×1027 atoms of 129I in the Arctic(35% of the total discharges from Sellafield and La Hague until 1996) a maximumtransport time of 11 years from the sources to the central Arctic is estimated.

Figure 5.4. The measureddistribution of 129I in majorocean layers in the centralArctic Ocean. Polar MixedLayer (PML), Atlantic Warmwater Layer (AWL), TransitionLayer (TL) and Deep OceanLayer (DOL), (paper III).

33

Table 5.1. Basic data and natural 129I/127I values in the central Arctic Ocean

LatitudeLongitude(deg.-min.)

Depth

(m)

129I/127I

(10-10)

06486-25.0144-29.6 59.8 35.8

249.9 16.5

497.5 14.0

815.3 10.8

862.7 10.3

06986-10.3155-00.3 61.1 41.1

250.4 13.1

600.4 9.6

1239.5 1.6

1287.7 1.5

08182-29.9134-36.3 20.7 57.7

274.6 9.8

1000.1 7.0

2999.6 0.3

4028.2 0.3

5.2 Precipitation

The main sources of natural 129I to the atmosphere are the oceans, through evaporation(e.g. Miyake & Tsunogai, 1963; Chamaides & Davis, 1980 and references therein;Whitehead, 1984). As described above, most of the 129I discharges from the Europeanreprocessing facilities are released into the western North Atlantic and have reached theArctic Ocean, which are important sources of precipitation to Scandinavia and Northern

34

Europe. Furthermore, direct discharges (between 2 and 20 % of the total liquid 129Idischarges) into the atmosphere have been documented (BNFL, 1998; GT1, 1999;UNESCAR, 1988). To gain some understanding of the atmospheric transport and falloutof 129I in Europe, an investigation on the 129I concentration in precipitation over centralSweden and northern Italy was carried out (paper IV). The results for the years 1998 and1999 show 129I concentrations at 4-40×108 atoms/liter in rain and at 0.4-3×108 atoms/literin snow over central Sweden, and at 0.3-9×108 atoms/liter in rain over northern Italy(paper IV). Correlation with factors such as wind direction, air temperature andchemistry of rain, for the day when the precipitation event occurred, was not strong (paperIV). Preliminary correlation with backward trajectories showed instead that the highest129I concentrations between September and October (figure 5.5) were measured when thesource of precipitation was located around Sellafield and La Hague. However, thespecific periodic enhancements in the 129I concentration could eventually also be relatedto an increase of deciduous material, such as leaves or organic residues from the soils(paper IV).

By using our data and an annual precipitation rate of 600 mm, we have estimatedthe flux of 129I from precipitation over Sweden (paper IV) to be 2×1023 atoms (0.043 kg)in 1998 and 4×1023 atoms (0.086 kg) in 1999. A similar estimation of the total depositionover Europe during 1998 can be done by considering an average of the 129I measured inprecipitation collected in other European countries (figure 5.6). The results show that in1998 the total 129I deposition amounted to about 1.3 kg, that represent about 85% of thetotal inventory of 129I (1.5 kg) in the atmosphere (ocean + land), as calculated for 1998 byusing the global model in chapter 2. Calculations performed using the model, withoutconsidering the atmospheric discharges lead to a 129I inventory in the atmosphere of only0.2 kg, suggesting that the main contribution of 129I to precipitation might be the gaseousdischarges into the atmosphere. This estimation seems to be consistent with resultsreported for rain over Germany, where the 129I/127I ratios is constant since 1987 (Szidat etal., 2000), reflecting more the behavior of the atmospheric rather than the one of theliquid discharges. However, this figure must be considered with precaution because it isbased on a global iodine evaporation rate of 1% of the total inventory in the surface oceanwhile, on a local scale, the evaporation rate could be higher. Further measurements andresearch are therefore necessary before a final conclusion can be reached.

35

Plotted together with other recent data of 129I in rain in Europe (figure 5.6), it isapparent that the highest concentrations (1010 atoms/liter) occur in northern Germany(Szidat et al., 2000). A simple explanation of this could be that the origin of theprecipitation is often located in the vicinity of the two point sources. The considerabledistance from the contaminated eastern north Atlantic is, instead, well reflected by thetwo orders of magnitude lower concentrations of 129I in precipitation over the U.S. (Moranet al., 1999).

Figure 5.6. 129I in rainwatersamples from differentlocations in Europe. G =Germany, Sp = Spain, Sw(Lopez-Gutiérrez, 1999) =Switzerland (Lopez- Gutiérrez,1999), S = Sweden, I = Italy(paper IV). G* is precipitationcollected in Germany belowtrees.

Figure 5.5. 129I concentrationin precipitation (filleddiamonds = rain and opensquares = snow) for 1998 and1999 plotted on a single yearx-axis starting from April (A)to May (M), showing a clearcycling change (paper IV).

106

107

108

109

1010

1011

U.S

.U

.S.

U.S

.Sp Sp Sw Sw G G

* S S It

129 I (

atom

s/lit

er)

1995

1996

1997

1996

1997

1996

1997

1996

/97

1996

/97

1998

1999

1998

107

108

109

1010

JM MJA J FA DOS N

129 I (

atom

s / li

ter)

36

5.3 Lakes, rivers and underground water

In order to evaluate sources of 129I to the inland surface water, a study of 129I in lakeslocated in Sweden and in Italy, in the river Fyrisån in Uppsala and in a few alpine creeksfrom the Alps was presented in paper V.

The lakes are located in central Sweden along a transect joining Stockholm toGävle, where the Chernobyl fallout was ranging from a few to more than 120 kBq/m2 for137Cs, offering a unique possibility of evaluating the contribution from the Chernobylsource. 129I concentration in the water of 15 lakes was measured. Even though flushingand renewal of the lake water have most likely removed 129I from the lakes 12 years afterthe accident, desorption of 129I from soil might have enhanced the 129I concentration,especially in those areas heavily contaminated by the accident. The results revealedinstead that the 129I concentration of the lakes was negatively correlated to the Chernobylfallout pattern, suggesting that 129I has mainly become fixed in the soil. Theconcentrations measured in the lakes are, instead, comparable to 129I concentrationsmeasured in precipitation (figure 5.7).

Figure 5.7. 129I measured in lakes of central Sweden in different months of the year 1998and ranges of rain and snow concentrations measured during 1998 and 1999 overSweden (papers IV and V). Cross = June, open squares = March, filled circles =November concentration. The numbers on the x-axis label the lakes according to paper V.

107

108

109

1010

1s 1d 2 3 4 5s 5d 6 7 8 9 10 11 12s

12d

13 14 15 16

129 I (

atom

s/lit

er)

highest rain concentration

lowest rain concentration

lowest snow concentration

37

Water from the river Fyrisån located in Uppsala shows values for the 129Iconcentration at 2-5×108 atoms/liter that are rather constant along the river course and thatreflect underground water contribution (paper V).

Lakes located in northern Italy (Pianura Padana), where the Chernobyl depositionwas much lower than in Scandinavia, have concentrations of 129I (ranging at 1-20×108

atoms/liter) similar to the ones measured in central Sweden (paper I). The tendency ofhaving slightly lower values, however, displayed by the lakes in Northern Italy comparedwith central Sweden, might reflect a barrier effect exerted by the Alps. The Europeanatmosphere seems, however, to be rather homogenized with respect to 129I, in accordancewith the short residence time (about 14 days) of iodine in the atmosphere. Small lakes orcreeks located in the Alps show concentrations of 129I of less than 107 atoms/liter (paperI), that mainly reflect contributions from groundwater.

38

CONCLUSIONS

In this thesis I have reached the following conclusions with respect to 129I analyticalmethods and results, and applications in natural water systems:

- Contribution of unwanted ions observed for 129I+7, 28 MeV, is totally eliminatedby using 129I+5, 21 MeV.

- The use of niobium as a binder, instead of silver, improves the detection limit of1 order of magnitude to the value of 10-14 through background reduction.

- A procedure for chemical extraction of iodine from water, allowing extendedinvestigations of a large number of samples (> 1000 per year), is developed. Asample to carrier ratio of 50ml/1mg can be used at 129I concentration > 108

atoms/liter.

- Storing samples (up to a period of 15 months) in dark and cold (4 ºC) roomwithout any treatment does not change the concentration of 129I.

- The central Arctic Ocean exhibits 129I concentrations in the order of 108-109

atoms/liter in the top layer and significant amounts are also measured in deeperparts (4000 m). The nuclear reprocessing facilities at Sellafield and La Hagueare the main sources of 129I in the Arctic Ocean through movement of watermasses. The estimated transit time of 129I from the sources to the central Arcticis less than 11 years.

- Fresh water environment in Europe, including precipitation, lake and runoffwater, has a 129I concentration of 108-109 atoms/liter. Lakes located in regionsheavily contaminated by the Chernobyl accident do not show exceptionalconcentration ranges. The 129I signal in precipitation seems to be related toatmospheric discharges from the nuclear reprocessing facilities in addition tothe likely release from the North Atlantic and the North Sea.

39

- The reprocessing releases from Sellafield and La Hauge are the major sourcesof 129I in the hydrological environment. Distinction between the role played bydifferent emissions (e.g. liquid and atmospheric discharges) is presently notfully evaluated. In order to quantify the impact of the atmospheric releases,continuos and systematic monitoring of 129I in precipitation, fresh water andmarine environment is needed.

40

APPENDIX. Data analysis

A.1. Measurement of the ratio of a sample

One cycle consists of one 129I measurement preceded and followed by a measurementof the current of the stable isotope 127I. After each cycle the ratio 129I/127I is calculatedas

i

ii I

CR =

Ci is the number of 129I ions counted at the detector and Ii is

( )2

III 1i1ii

+− +=

where Ii-1 and Ii+1 are the 127I currents measured at the Faraday cup after the analyzingmagnet before and after the 129I measurement, respectively.

The error (dRi) on the ratio Ri is assumed to be due to statistical errors associated with129I measurements, so that

i

ii I

CdR =

When “n” cycles of measurements are performed, the best value for R is given by:

=

== n

1i2

i

n

1i2

i

i

)dR(1

)dR(R

R

the Poisson error is given by

=

= n

1i i

poiss

dR1

1dR

41

and the standard deviation by

=

=

−

−

= n

1i2

i

n

1i2

i

2i

sd

)dR(1)1n(

)dR()RR(

dR

A.2. Normalization to the standard

The ratio of the standard and its error are calculated as presented above. Each sampleratio is normalized to the ratio of the standard. A standard is measured after every 4unknown samples. The standard ratio value used for normalizing each unknownsample is the result of a linear interpolation of the two measurements (before and afterthe 4 unknown samples) of the standard in function of the time.

The ratio (S) between the sample (Rsam) and the standard (Rst) is

st

sam

RRS =

and the error is found by applying the propagation error rule

2

st

st

2

sam

sam

st

sam

RdR

RdR

RRdS

+

=

When the sample is measured more than one time (n sequences), the weighed mean ofthe “n” ratios is calculated as

=

=

=

n

1i2

i

n

1i2

iist

sam

st

sam

)dS(1

)dS(1

RR

RR

with errors given by

42

=

=

n

1i2

i

st

samPoiss

)dS(1

1R

RdS and 2

i

2i

2i

st

samst

dS1

)dS()SS(

RRdS

−

=

From the value of the ratio

st

samp

RR

, it is possible to calculate the real value of

the 129I/127I ratio in the sample by multiplying

st

samp

RR

by the nominal value of the

ratio in the standard, that in most cases was 1.174×10-11.

A.3. Calculation of the numbers of 129I atoms in the sample

The number of 129I in atoms per liter of water is calculated according to the followingformula

sample127

12923129

II

)liter(Volume)liter/g(Carrier

)mol/g(945.126)mol/atoms(10022.6)liter/atoms(I

×=

The error is calculated by considering the measurement error on the 129I/127I ratio of thesample as

2129

RdR)liter/atoms(Ierror

=

A.4. The background correction

In AMS measurements, a background correction is necessary because the samplesusually acquire an amount of contamination either from the environment or during thechemical treatment. Instrumental background is also present but is much less than thecontamination mentioned above. In order to be able to make a background correction,“blank” targets are chemically prepared in the same way as the samples. Themeasured isotopic ratio of a blank provides the background limit and isotopicsensitivity of the AMS measurement and can be expressed in the case of iodine as

43

cstd

cstd

cb

c

std

b

)127()127()129()129()127()127(

)129(

)127/129()127/129(f

+++==

where the subscripts b, std and c symbolize blank, standard and contamination. Byexpressing the measured ratio Fm of the samples as

cstd

cstd

cs

cs

std

sm

)127()127()129()129()127()127()129()129(

)127/129()127/129(F

++++

==

where s signifies sample, the real value of the ratio F in the sample is derived by thefollowing equality (Winkler et al., 1997)

f1fFF m

−−=

If f << 1 then the background correction is simply done by subtracting the blank fromthe measured ratio.

Table A.1. Values of f (together with standard deviation from the mean of the valuesobtained after many measurements) and 129I/127I for different carriers. Woodwardiodine is extracted from a deep underground brine source 300 million years old(Woodward Corporation, Oklahoma) while the old Precambrian (> 300 Ma old)natural crystal is iodargyrite

f 129I/127I(×10-13)KI 0.02±0.01 2AgI (from KI) 0.06±0.008 2 – 20Woodward 0.02±0.003 0.4 – 3Iodargyrite 0.008±0.002 0.5 – 2

Results of numerous measurements of blank targets during the last 3 years showdifferent values of f depending on the different kinds of carrier used to prepare the

44

blank targets. The average values of f and the ranges of 129I/127I ratios of a super-purepotassium iodide (KI), AgI targets prepared from KI, Woodward iodine (I2), and anold natural crystal (AgI) measured in Uppsala are summarized in table A.1.

45

References

Baetsle L. H., Impact of fission product partitioning and transmutation of 237Np, 129Iand 99Tc on waste disposal strategies, Belgian Nuclear Research Centre,Proceedings of the First International Information Exchange Meeting onActinide and Fission Product Separation and Transmutation, Mito, Japan,1990.

Beasley T. M., Cooper L. W., Grebmeier J. M. & Kilius L. R., 129I in western Arcticwaters, in Environmental Radioactivity in the Arctic, edited by P. Strand and A.Cooke, 121, 1995.

Betz H.-D., Charge states and charge-changing cross sections of fast heavy ionspenetrating through gaseous and solid media, Reviews of Modern Physics, 44, 3,1972.

BNFL, Annual report on discharges and monitoring of the environment, 1998.

Boaretto E., Berkovits D., Hollos G. & Paul M., Positive identification of XeH-

molecular ion, Nuclear Instruments and Methods in Physics Research, B 52,421, 1990.

Busch F., Pfeffer W., Kohlmeyer B., Schüll D. & Pühlhoffer F., A position sensitivetransmission time detector, Nuclear Instruments and Methods in PhysicsResearch, B 171, 71, 1980.

Chameides W. L. & Davis D. D., Iodine: its possible role in troposphericphotochemistry, Journal of Geophysical Research, 85, C12, 7383, 1980.

Dean G. A., The iodine content of some New Zealand drinking waters with a note onthe contribution from sea spray to the iodine in rain, New Zealand Journal ofScience, 6, 208, 1963.

Edmonds H. N., Smith J. N., Livingston H. D., Kilius L. R. & Edmond J. M., 129I inarchived seawater samples, Deep-Sea Research I, 45, 1111, 1998.

Elmore D., Gove H. E., Ferraro R., Kilius L. R.,. Lee H. W, Chang K. H., Beukens R.P., Litherland A. E., Russo C. J., Purser K. H., Murrell M. T. & Finkel R. C.,

46

Determination of 129I using tandem accelerator mass spectrometry, Nature, 286,138, 1980.

Fabryka-Martin J. T., Natural 129I as a ground-water tracer, Master of Science thesis,University of Arizona, 1984.

Fehn U., Holdren G. R., Elmore D., Teng R., Tullai S., Kubik P. W. & Gove H. E.,Measurement of 129I in the marine environment, Terra Cognita 6, 1986.

Fuge R. & Johnson C., The geochemistry of iodine - a review, EnvironmentalGeochemistry and Health, 8(2), 31, 1986.

Fuge R., Geochemistry of iodine in relation to iodine deficiency diseases, inEnvironmental Geochemistry and Health, Geological Society SpecialPublication, Appleton, Fuge J. D. & McCall G. J. H. editors, 113, 201, 1996.

GT1, Groupe Radioecologie Nord-Cotentin, Inventaire des rejets radioactifs desinstallations nucléaires, 1, 1999.

Hilmer, M., Harder M. & Lemke P., Sea ice transport: a high variable link betweenArctic and North Atlantic, Geophysical Research Letters, 25, 17, 3359, 1998.

Hou X., Dahlgaard H., Rietz B., Jacobsen U., Nielsen S. P. & Aarkrog A.,Determination of chemical species of iodine in seawater by radiochemicalneutron activation analysis combined with ion-exchange preseparation,Analytical Chemistry, 71, 14, 2745, 1999.

Hou X., Dahlgaard H. & Nielsen S. P., 129I time series in Danish, Norwegian andnorthwest Greenland coast and the Baltic Sea by seaweeds, Estuarine Coastaland Shelf Science, in press, 2000.

Jones S. D., Studies of speciations of iodine in rain and freshwater, Tesis Doctoral,University of Wales, Aberyswyth, 1981.

Jones S. D. & Truesdale V. W., Dissolved iodine species in a British freshwatersystem, Limnol. Oceanogr., 29, 1016, 1984.

Josefsson D., Holm E., Persson B. R., Roos P., Smith J. N. & Kilius L., Radiocaesiumand 129I along the Russian coast. Preliminary results from the Swedish Russian

47

Tundra Ecology expedition 1994, in Environmental Radioactivity in the Arctic,edited by P. Strand and A. Cooke, 273, 1995.

Josefsson D., Anthropogenic radionuclides in the Arctic Ocean, distribution andpathways, ISBN 91-628-2967-X, Ph.D. thesis, Lund University, 1998.

Kaplan, D. I., Serne R. J., Parker, K. E. & Kutnyakov, I. V. Iodine sorption tosubsurface sediments and illitic minerals, Environmental Science &Technology, 34, 399, 2000.

Kilius, L. R., Zhao X.-L., Smith J. N. & Ellis K. M., The measurements of 129I in theCanadian Arctic Basin and other Arctic waters, in Environmental Radioactivityin the Arctic, edited by P. Strand and A. Cooke, 117, 1995.

Kilius L. R., Kieser W. E., Litherland A. E., Zhao X.-L., Rucklidge J. C. & BeukensR. P., Ion source design criteria for AMS, Nuclear Instruments and Methods inPhysics Research, B 123, 5, 1997a.

Kilius L. R., Zhao X.-L., Litherland A. E. & Purser K. H., Molecular fragmentsproblems in heavy element AMS, Nuclear Instruments and Methods in PhysicsResearch, B 123, 10, 1997b.

Kocher D. C., A dynamic model of the global iodine cycle and estimation of dose tothe world population from releases of 129I to the environment, EnvironmentalInternational, 5, 15, 1981.

König M., AMS measurements of 129I at the Uppsala tandem laboratory, Lic. Thesis,UPTEC 97 101R, 1997.

Liu Y. & von Guten H. R., Migration chemistry and behaviour of iodine relevant togeological disposal of radioactive wastes - a literature review with acompilation of sorption data, Technischer Bericht 88-29, 1988.

López-Gutiérrez J. M., Aplicaciones de la Espectrometria de Masas con Aceleradoresa la determinación de radionúclidos de semivida grande en la naturaleza, TesisDoctoral, Sevilla, 1999.

Luten J. B., Woittiez J. R. W., Das H. A. & De Ligny C. L., Determination of iodate inrain-water, Journal of Radioanalytical Chemistry, 43, 175, 1978.

48

Middleton R., A review of ion sources for accelerator mass spectrometry, NuclearInstrument and Methods in Physics Research, B5, 193, 1984.

Miyake Y. & Tsunogai S., Evaporation of iodine from the ocean, Journal ofGeophysical Research, 68, 13, 3989, 1963.

Moran J. E., Teng R. T. D., Rao U. & Fehn U., Detection of iodide in geologicalmaterials by high-performance liquid chromatography, Journal ofChromatography A, 706, 215, 1995.

Moran J. E., Oktay S., Santschi P. H. & Schink D. R., Atmospheric dispersal of 129Ifrom nuclear fuel reprocessing facilities, Environmental Science & Technology,33, 2536, 1999.

Moran S. B., Cochran J. K., Fisher N. S. & Kilius L. R., 129I in the Ob river, inEnvironmental Radioactivity in the Arctic, edited by P. Strand and A. Cooke,75, 1995.

Purser K. H., Litherland A. E. & Gove H. E., Ultra-sensitive particle identificationsystems based upon electrostatic accelerators, Nuclear Instrument and Methodsin Physics Research, B 162, 637, 1979.

Raisbeck, G. M., Yiou F., Zhou Z. Q., Kilius L. R. & Dahlgaard H., Anthropogenic129I in the Kara Sea, in Environmental radioactivity in the Arctic and Antarctic,edited by P. Strand and E. Holms, 125, 1993.

Raisbeck G. M., Yiou F., Zhou Z. Q. & Kilius L. R., 129I as a tracer of reprocessingdischarges in the Arctic Oceans, in Environmental Radioactivity in the Arctic,edited by P. Strand and A. Cooke, 109, 1995.

Raisbeck G. M. & Yiou F., 129I in the oceans: origins and applications, The Science ofthe Total Environment, 237/238, 31, 1999.

Rahn K. A, Borys R. D. & Duce R. A., Tropospheric halogen gases: inorganic andorganic components, Science, 192, 549, 1976.

Rao U. & Fehn U., Sources and reservoirs of anthropogenic iodine-129 in westernNew York, Geochimica and Cosmochimica Acta, 63, 13/14, 1927, 1999.

49

Roberts, M. L., Caffee, M. W. & Proctor, I. D., 129I Interlaboratory Comparison: PhaseI and phase II results. ISPO-394, UCRL-ID-128218, 1997.

Smith J. N., Ellis K. M. & Jones E. P., Cesium 137 transport into the Arctic Oceanthrough Fram Strait, Journal of Geophysical Reasearch, 95, C2, 1693, 1990.

Smith J. N., Ellis K. M. & Kilius L. R., 129I and 137Cs tracer measurements in theArctic Ocean, Deep-Sea Research I 45, 959, 1998.

Strand P. & Holm E. editors, in Environmental Radioactivity in the Arctic andAntartic, 19, 1993.

Studier M. H., Postmus C. Jr., Mech J., Walters R. R. & Sloth E. N., The use of 129I asan isotopic tracer and its determination along with normal 127I by neutronactivation - the isolation of iodine from a variety of materials, Journal ofInorganic and Nuclear Chemistry, 24, 755, 1962.

Szidat S., Schmidt A., Handl J., Jakob D., Botsch W., Michel R., Synal H.-A., SuterM., Lopéz-Gutiérrez J. M. & Städe W., Iodine-129: sample preparation, qualitycontrol and analyses of pre-nuclear material and of natural waters from LowerSaxony, Germany, Nuclear Instrument and Methods in Physics Research B,AMS-8 conference proceedings, in press, 2000.

Takayanagi K. & Wong G. T. F., The oxidation of iodide to iodate for thepolarographic determination of total iodine in natural waters, Talanta, 33, 5,451, 1986.

Tsunogai S. & Sase T., Formation of iodide-iodine in the ocean, Deep Sea Research,16, 489, 1969.

UNSCEAR. Sources, Effects and Risks of Atomic Radiation; U.N. Committee on theEffects of Atomic Radiation: New York, 1988.

Yiou F., Raisbeck G. M., Zhou Z. Q. & Kilius L.R., 129I from nuclear fuelreprocessing; potential as an oceanographic tracer, Nuclear Instruments andMethods in Physics Research, B 92, 436, 1994.

Yiou F., Raisbeck G. M., Zhou Z. Q., Kilius L. R. & Kershaw P. J., Improvedestimates of oceanic discharges of 129I from Sellafield and La Hague, in

50

Environmental Radioactivity in the Arctic, edited by P. Strand and A. Cooke,113, 1995.

Wagner M. J .M., Dittrich-Hannen B., Synal H.-A., Suter M. & Schotterer U., Increaseof 129I in the environment, Nuclear Instruments and Methods in PhysicsResearch, B 113, 490, 1996.

Whitehead D. C., The distributions and transformations of iodine in the environment,Environment International, 10, 321, 1984.

Winkler G., Priller A. & Hille P., On the background correction for AMS 14Cmeasurements, http://www.univie.ac.at/Kernphysik/research/ams.htm, 1997.

Wong G. T. F., The marine geochemistry of iodine, Reviews in Aquatic Sciences, 4(1),45, 1991.

Åström J. & Possnert G., The Uppsala cesium sputter ion source, internal report TLU102, 1983.

51

Acknowledgments

A Ph.D. study, especially when carried out abroad, represents a journey when theprocess of learning several aspects of science and of life is of inestimable value. Thereare many people that played a role in this part of my life. Persons working with mebecause they have put effort and care into my project. And friends because I haveshared with them moments of happiness.

My deepest gratitude goes to my supervisor, Göran Possnert, because he guided meprofessionally and with kindness during all these years. I thank him for what he taughtme about AMS, for always having assisted me during the evening-time acceleratorruns, for the long working meetings, and for having answered all my questions, evenwhen he was extremely busy. I appreciated the opportunity of working with himbecause from his special, pleasant and skilled personality I have realized and learntmany things that I will carry with me during all my life.

Particular thanks are due to Ala Aldahan, my assistant supervisor, who gave aninestimable contribution to my work. He introduced me to the world of chemistry,geology and environmental science but also he taught me how to write papers. Hiskind, cheerful and caring attitude inspired me a lot and encouraged me to go on, evenwhen things felt a little bit challenging.

I am grateful to Rector Magnificus Prof. Bo Sundqvist, who was responsible for theIon Physics Group when I came to Uppsala and who gave me the unique possibility tostart working in the group. I am also grateful to Per Håkansson because he gave meuseful advice on several occasions.

If I now am able to measure 129I at the tandem, it is in the main because BengtBeckman, Cliff Robinson (who is also a good personal friend of mine) and RogerioZorro spent part of their time teaching me with patience how to run an accelerator. Ihave enjoyed the time spent with them because of the nice working atmosphere andbecause of the numerous chats that made my iodine runs more pleasant.

During the beginning of my time here I have shared the office with Martin König, whowas working on iodine at that time. Thanks for the discussions about 129I and foralways creating a friendly atmosphere in the group by organizing beer evenings,dinners, parties and several kinds of games.

My gratitude goes to Inger Påhlsson for assisting me during first tests of chemicalextractions of iodine and for all the invaluable help received from her in the chemistry

52

laboratory. I am also grateful to Markus Meili for the long and interesting discussionsabout biochemistry.

Thanks to Paula Almqvist, our nice and efficient secretary, because she alwaysanswered my numerous questions and helped me fill in some of the documents thatwere almost incomprehensible to me. Thanks also for the well arranged dinners andhorse riding excursions.

I also want to express my gratitude to my fellow graduate students and posdocs in theIon Physics Group: Udaya Abeywarna, Vassili Alfimov, Jan Axelsson, Edy Ayala, JosBuijs, César Costa Vera, Mikael Danfelter, Jan Eriksson, Charlotte Hagman, JanHinnerk Richter, Kristina Håkansson, Judit Kopniczky, Srimalie Nugegoda, MagnusPalmblad, Ricardo Papaléo, Gamini Piyadasa, Margareta Ramström, Youri Tsybin,Bernd Umann and all the German exchange students for great moments both insideand outside the lab.

Maud Söderman and Britt-Marie Ström for being nice and patient with me when Ispeak Swedish, Hjördis Andersson, Eva-Lotta Anlér-Blomberg, Raul Figueroa,Thomas Hammar, Tomas Kronberg, Johan Kjellberg, Yngve Kjellgren, Gösta Widmanand Jonas Åström for the nice atmosphere created during the coffee-break.

Thanks to all the nice friends I met in Sweden: Anna for always listening to me andmy “long stories”, for the wonderful days spent in Kasmas, for being so nice andhospitable; Aziz for his real friendship, for all our talks and for his Italian way ofbeing; Curt for cooking such good food, for organizing that wonderful party at mybirthday, for his jokes and for his comments on how to write in English. Also, Ulrika,Raad, Simone, Annarita and Yanwen for the nice and relaxing moments spent together.Simone and Filippo for cheering me up every time I talk to them, Carl for his long e-mails, Alain for his Swiss chocolate, Luigi, Paolo, Arjan, Elisabeth, Hadi, Johannes,Stephan, Hendrik and Menhel for their care and friendship.

In these years, I also enjoyed practicing sports. Thanks to all the Tennis-people(especially to my German trainer Mau), and to the Bandy group for many matchesplayed together and to the members of USDK and CAMBALACHE dance club.

I also want to remember my Italian friends because they have always stayed close tome even when I moved to Sweden. Their support was extremely valuable especially inthe beginning. Among them I will mention my best friend Katia, Carlo and Sonia forcalling me almost every week, Marco, Barbara, Moreno, and all the members of thetheatre group.

53

I wish to thank my boyfriend’s parents, Sandra and Pietro, for having alwaysencouraged us in our studies, for having left us a car that made our life (especiallymine) much easier, for the vacations spent together and the “dancing evenings” here inSweden.

My wonderful parents, Luigi and Elena, for having always supported and encouragedme, and my beloved and sweet brothers, Roberto and Carlo, for loving me andkeeping me always updated with any news about home. To my grandmothers Onorinaand Fernanda because I have partially inherited their fantastic way of being. And toShiro.

Finally, I want to thank “my unique and beloved Marco” who has been the initiator ofour adventure in Sweden. Thanks for your care, for always listening to me withpatience, for all the support and for your love. Your presence was essential to go on inthe most difficult moments and to deeply enjoy the most wonderful ones.

Uppsala, September 2000