Nuclear fingerprinting of radioactive material in a radiological and nuclear 

safeguard perspectiveMats Eriksson

SSM

Outline

What is nuclear fingerprinting? Safeguard and radiological perspective Bulk versus particle analysis (why study particles) Overview of Fingerprinting techniquesLocalisation techniques Radiometric techniquesX‐ray techniques  Mass spectrometric techniques

Conclusions Wake you up

Nuclear fingerprintor

Nuclear material characterisation 

Trace of the source

In Safeguards the Uranium and Plutonium isotopic composition are the focus to check that member states follow the Non‐Proliferation Treaty (NPT). 

The Safeguard perspective 

With Radiological studies** it is not only the elemental and isotopic ratios composition that is of interest, but also the chemical form and radionuclide distribution. This information is essential for performing adequate dose assessments and for predicting the long term effects of radioactive contamination in an area.

The radiological perspective 

** (means both the radio-ecological and emergency preparedness studies in this talk)

U and Pu, the most studied elements

Different categories of U and PuCategory 235U/238U(%) 240Pu/239Pu (%)Depleted uranium (DU) < 0.71Natural uranium 0.71Low enriched uranium (LEU) 0.71-20 (usually 3-5)Highly enriched uranium (HEU) > 20 (usually 90)Weapons-grade uranium (WGU) > 90Reactor-grade plutonium (RGPu) > 19Fuel-grade plutonium (FGPu) 7-19Weapon-grade plutonium (WGPu) < 7Super-grade Pu (SGPu) < 3

Plutonium fingerprints from different sources

Plutonium isotope atom ratios for some Pu sources 238Pu/239Pu 240Pu/239Pu 241Pu/239Pu 242Pu/239Pu

Atmospheric fallout (17.7±0.32)·10-4 0.1808±0.0057 (2.64±0.20)·10-3 (4.0±0.3)·10-3

(Northern hemisphere average)

Kyshtym, accidental release (3.34±0.89)·10-5 0.0282±0.0001 (2.31±0.06)·10-4 (7.15±0.26)·10-5

Sellafield discharge 0.0118 0.1838 0.0116 0.0053Chernobyl accident 0.00334 0.563 0.14 0.0429Thule accident (3.35-5.9)·10-5 0.027-0.058 (0.3-4.5)·10-4

Data extracted from Cooper et al. and Eriksson et al, 2008 (decay corrected to January 1, 1995).

Uranium isotopic compositions from different nuclear facilities exist in a database at the safeguard department.

Studies on the isotopic composition can be used to identify the source

Bulk versus particle analysis

Safeguard perspective It is the particles originating from the nuclear facility that are the fingerprints of the facility’s activities ‐ bulk material will blur and bias the information.

Radiological perspective In nuclear events involving explosions or fire, most of the material will be released as particles.

If several sources are present, the bulk sample will show the mean of the sources. 

Environmental parameters can be obtained easier  from particles than from bulk samples

Log normal distribution ‐ Size converted to Mass /Activity distribution; A few large sized particles carry the majority of the mass/ activity released.

Examples of log normal distributionsMaterial dispersed after an explosionMineral resources in the earth’s crustPollutants in the air 

Log‐normal particle size distributionHeterogeneous activity distribution

Two ways of characterizing lognormal distributions, in terms of the original data (a) and after log-transformation (b).

http://www.inf.ethz.ch/personal/gut/lognormal/brochure.html

1.3 % of the particles carry 1.3 % of the particles carry 80% of the activity80% of the activity

1 Bq Pu particle ≈ diameter of 20µm

Particle activity distribution, small scale spatial variance, effect on the inventory calculation

0 2 4 6 8 10 12

02

46

810

12

particles, black=1,red=10, green=100

0 2 4 6 8 10 12

02

46

810

12

particles, black=1,red=10, green=100 Log normal distribution

Assuming homogenous dist.:∑=500 units 196 cells gives:Average unit per cell = 2.55Average unit per row = 35.7However, about 50% of the cells contain no particles = 0 units

Particles;  ‐ 100 black a’ 1 unit ‐ 10 red a’ 10 units ‐ 3 green a’ 100 units

Obtaining representative, unbiased samples Usually activity concentrations are based on the assumption of a homogenous activity distribution and are calculated from a few sub samples• Sampling with small samples ‐ spatial activity distribution will lose 

meaning • Sampling with large samples ‐may hide meaningful variations

Formed in events involving explosions (e.g. the Chernobyl and the Thule accidents) and at nuclear installations (power and reprocessing plants)

The fingerprint can reveal the source terms  They “carry” the main portion of the mass 

released, however they are very rare, leading to a heterogeneous activity distribution

Mostly a close‐in fallout problem  “Contain” geochemical “memory” Potential radiation hazard as they may lead to 

a high local absorbed dose

Why study particles

R.Pöllänen. Ph.D thesis, 2002

Safeguards Sampling work

Sampling in a radiological studySediment sampling grid during Thule‐97

Large scale, cover all the contaminated sediments

Gemini sediment sampler

Gemini slicing

Finding and Identification Image techniques (e.g. image plates, beta camera, 

SEM,…) Analytical techniques Elemental and isotopic techniques: 

SEM‐EDX Radiometric techniques: Gamma and alpha spectrometry Mass‐spectrometric techniques: ICP‐MS and SIMS

Physicochemical characterisations   Synchrotron radiation techniques and PIXE

µ‐XRF, µ‐XRF tomography and µ‐XANES

Isotopic fingerprinting techniques applied to particles

Pu Hot Particle

Particle separation and localization techniques

20 h acquisition time 1 h acquisition time

Sampling splitting

Fixation on adhesive carbon tape

Measurements for identificationIDE Bioscope Beta camera

241PuT½=14.4y

241AmT½=432y

237NpT½=2.1 106 y

α

β-

γ : 59.5 keV

241PuT½=14.4y

241AmT½=432y

237NpT½=2.1 106 y

α

β-

γ : 59.5 keV

Tritium imageplates

1 week exposure time

Impactor and SIMS for U particles

Fingerprint indicatorsRadionuclide composition Activity determinationInformation on density and shape can be gained

Radiometric techniques 

Gamma- alpha- coincidence detector system in list mode

Gamma spectrometryHigh active sample

Figure from Ulrika Nygrens Ph.D. 2006

Non-destructive analysis238Pu239Pu240Pu241Pu

241Am

Characteristic L‐x‐ray, 241Am(241Pu) to Pu

Relatively Low 241Am/238+239+240Pu 0.13

Relatively High 241Am/ 238+239+240Pu 0.17

P = n A solution:n-1 P = A

Gamma spectrometryLow active sample

Direct Alpha spectrometry and fitting program

AASI, shape, density, activityhttp://www.stuk.fi/tutkimus/programs/en_GB/aasi/

239+240Pu

Two 238Pu/239,240Pu ratio groupsAverage ratio: 0.014±0.004 (n=328)

238Pu242Pu

SEM-EDX: Morphology, size and elemental compositionMapping & Point measurements

U/Pu Lαratios

Example of heterogeneous and homogenous U and Pu distributions in particles

U/Pu Lα ratios

EDX Spectra

XRF experimental set‐up at the synchrotron radiation (SR) facilities

From 2D to 3D elemental information Micro‐tomography experimental set‐up (combined 

transmission and XRF‐tomography) Micro XANES (speciation of elements)

Micro PIXE Confocal XRF

The X‐ray techniques applied in particle characterisations

Synchrotron Radiation base ‐ Fluorescence Spectroscopy

X‐ray Fluorescence (XRF) Spectroscopy elemental mapping (micro‐focused beam size mode, micro‐XRF)

AdvantagesMono energetic and high flux beamNon destructive, surface / volume sensitiveSpectroscopy from light elements (Al) to Am with a single beamline

Micro XRF and confocal XRF setup at Beam Line L in HASYLAB, Germany

Confocal setup

DESY

From 2D to 3D with Confocal µ‐XRF (add. Information)

Synchrotron 3D XRF setup at ANKA

The IAEA set‐up at the synchrotron beamline

X‐ray Fluorescent computer tomography (XFCT)

SEM image Density imageof one slice Elemental reconstruction

Elemental distribution, µ‐ XRF Tomography

Thule particle; Pu (blue), U (green) and Fe (red)

Sellafield particleCa (blue-green), U (yellow)

Mururoa particle; Pu (blue), Sr(yellow)

Mental “bensträckare” slide !Showing other examples of micro tomography

Sterile mosquito project, morphology changes?Wegrzynek et al, 2005 ICXOM

300 m

Fe As3+

As5+

Arsenic Distribution in Cattail Roots, Nicole Keon, Harold Hemond, Daniel Brabander (MIT):

Determination of the elemental distribution in human joint bones by SR micro XRF, Zoeger et al 2008, X-Ray Spectrom. 2008; 37: 3–11

QUANTITATIVE FUNCTIONAL IMAGING AND KINETIC STUDIESWITH HIGH-Z CONTRAST AGENTS USING SYNCHROTRONRADIATION COMPUTED TOMOGRAPHY, Adam et al, Clinical and Experimental Pharmacology and Physiology (2009) 36, 95–106

µ‐XRF vs. µ‐XRF Confocal

Particle ID Pu/U ratio (voxel derived) ( 1 SD)

Relative uncertainty

Pu/U ratio (summed spectra ) (1 SD)

Thu 68-1 0.20 ± 0.02 ( n= 67 ) 10 0.172 ± 0.005

Thu 97-1 N.A. N.A 0.20 ± 0.01

Thu 975371-4 0.12 ± 0.03 ( n= 622 ) 25 0.113 ± 0.008

Thu 2003-7524 0.38 ± 0.33 ( n= 414 ) 87 0.235 ± 0.002

T2003-7524, Pu/U Lalfa ratio mean=0.43 +/- 0.38 (1 sd)

mean=0.28 +/- 0.07 (ex ratio >0.5)

Pu/U Lalfa ratio

Freq

uenc

y

0.0 0.5 1.0 1.5 2.0 2.5

020

4060

8010

0

7

27

88

98

61

38

119

56 6 62

6

132 3 2 1

5

0

51

5

0 00 032 1 0 0

20 12

0 1 01 200 0 00 0 1

Preferential U leaching

Micro‐PIXE a complementary technique to µ‐XRF

Very good correlation with confocal XRF

Absorption Spectroscopy

information about the local atomic geometry (EXAFS) chemical state of the absorbing atom (XANES)

Lα ROI intensity

Lα ROI

Chemical speciation; Pu and U XANES

Both U and Pu in Thule particlesare mostly in the +4 oxidation state, i.e. a low solubility phase.

Mass‐spectrometric techniques 

Higher sensitivity to long lived radionuclides than to radiometric techniques

Isotopic ratios Mass spectra Depth profile (SIMS)

ICP-MS on bulk soil samples24

0 Pu/

239 P

u m

ass r

atio

Pu isotopic ratio results from Thule soil samples

In soil contaminated with U material it is very difficult to find the source U fingerprint from bulk soil sample analysis, as natural U is present everywhere causing a bias in the fingerprint.

The SIMS technique

SIMS Results-Isotopic Ratio

Thu 79-2

234U/235U 236U/235U 235U/238U 240Pu/239Pu 242Pu/239Pu (241Pu+241Am)/239Pu

Thule68−1 0.0108±0.0002

0.0090± 0.0004 1.355± 0.006 0.0578± 0.0005 0.00018± 0.00002 0.0082± 0.0002

Thule68−2 0.0108±0.0001

0.0129± 0.0003 0.893± 0.002 0.0457± 0.0004 0.00011± 0.00003 0.0034± 0.0001

Thule68−3 0.0111±0.0003

0.0142± 0.0004 1.199± 0.006 0.0580± 0.0006 0.00028± 0.00003 0.0099± 0.0002

Thule79−4 0.0107±0.0002

0.0048± 0.0002 1.440± 0.006 0.0364± 0.0004 0.00008± 0.00002 0.0052± 0.0002

Thule79−5 0.0112±0.0004

0.0104± 0.0003 1.302± 0.010 0.0631± 0.0008 0.00045± 0.00006 0.0062± 0.0002

Thule79−6 0.0107±0.0008

0.0103± 0.004 1.032± 0.012 0.0284± 0.0004 0.00003± 0.00002 0.0024± 0.0001

Thu975380−5nr1 0.0109±0.0003

0.0102± 0.001 1.376± 0.009 0.0580± 0.0007 0.00017± 0.00004 0.0106± 0.0003

Ratio range 0.0107-0.011 0.0048-0.0129 0.893-1.440 0.0364-0.0631 0.00003-0.00045 0.0034-0.0106

WGU WGPu

Nuclear fingerprints of the Thule accident

The different ratios indicate that the particles originated from different weapon components involved in the Thule accident

SIMS depth profile, preferential leaching?

10 µm

Combined techniques: micro‐XRF and SIMS 

Preferential leaching of U compare to Pu?

SEM image µ-XRF map of U and Pu U/Pu from µ-XRF with the SIMS depth profile superimposed

A new source in Thule ?

20 40 60 80 100 120

2040

6080

100

x

y

20 40 60 80 100 120

2040

6080

100

x

y

High 235U enriched particles235U /238U = 8.2 ± 0.1

All U voxels

log((U[log(U) > -2]))Fr

eque

ncy

-2 0 2 4

050

010

0015

0020

0025

00

U with Pu voxels

log((U[Pu > 0.05 & log(U) > -2]))

Freq

uenc

y

-2 0 2 4

020

040

060

080

0

Pure U voxels

log(RenU[log(RenU) > -2])

Freq

uenc

y

-2 -1 0 1 2 3

050

010

0015

00

All voxels

log((U[log(U) > -2]))

Freq

uenc

y

-2 0 2 4

050

010

0015

0020

0025

00

All U voxels

log((U[log(U) > -2]))Fr

eque

ncy

-2 0 2 4

050

010

0015

0020

0025

00

U with Pu voxels

log((U[Pu > 0.05 & log(U) > -2]))

Freq

uenc

y

-2 0 2 4

020

040

060

080

0

Pure U voxels

log(RenU[log(RenU) > -2])

Freq

uenc

y

-2 -1 0 1 2 3

050

010

0015

00

All voxels

log((U[log(U) > -2]))

Freq

uenc

y

-2 0 2 4

050

010

0015

0020

0025

00

Fig. 3. This figure shows the U (red-yellow pixel image) and the Pu (the superimposed iso-intensity lines) distribution in one cross-section of the particle shown in Fig 2. A line profile of the Pu/U ratio is also shown (blue line in the figure). The Pu/U ratio is multiplied by a factor of 100.

B

A

C0.88 eV1.88 eV1.74 eV

Rel Pu L3 Shift to Pu(+3)

CBA

Area

µ-XANES

0 1000 2000 3000 4000 5000 6000Time [s]

0.3

0.4

0.5

0.6

0.7

0.8

5.4

5.5

5.6

5.7

5.8

1.4

1.6

1.8

2.0

Fingerprints obtained through Combined techniques

SIMS depth profile

Age determined from 241Am/241Pu in growth analysis‐Weapon material produced: 1958 +/‐ 5y 

Conclusions

• Insight into Micro analytical techniques and how they can be applied in the Safeguard and radiological fields to reveal nuclear fingerprints

• Particles are suitable for revealing nuclear fingerprints

• Particles are important to consider when undertaking radiological studies

More Specific Conclusions

Particles are important:

• Carry most of the released material after an explosion and need to be taken into account in inventory estimates

• Can reveal the source term of an event • Can be used as geochemical “monitors” eg in spent fuel

dissolution studies as “natural” analogues• Radiation point sources with a potential to give a high local

absorbed dose• Essential in nuclear safeguard analysis

Institute for Transuranium ElementsYlva RaneboG. Tamborini

Acknowledgements to:Risø National Laboratory

Henning DahlgaardJussi Jernström

CNA PIXE team, SevillaM.-C. Jimenez Ramos

R. Garcia-TenorioJ. Garcia Lopez

SR Beam-line scientists:R. Simon (ANKA)

G. Falkenberg (HASYLAB)

IAEA :D. Wegrzymek

O. BildsteinD. DonahuM. Betti

STUKR. Pöllänen