This content has been downloaded from IOPscience. Please scroll down to see the full text.

Download details:

IP Address: 113.97.192.243

This content was downloaded on 17/10/2013 at 11:44

Please note that terms and conditions apply.

Gamma-ray imaging of D and 4He ions accelerated by ion-cyclotron-resonance heating in

JET plasmas

View the table of contents for this issue, or go to the journal homepage for more

2005 Nucl. Fusion 45 L21

(http://iopscience.iop.org/0029-5515/45/5/L01)

Home Search Collections Journals About Contact us My IOPscience

INSTITUTE OF PHYSICS PUBLISHING and INTERNATIONAL ATOMIC ENERGY AGENCY NUCLEAR FUSION

Nucl. Fusion 45 (2005) L21–L25 doi:10.1088/0029-5515/45/5/L01

LETTER

Gamma-ray imaging of D and 4He ionsaccelerated by ion-cyclotron-resonanceheating in JET plasmasV.G. Kiptily1, J.M. Adams1, L. Bertalot2, A. Murari3,S.E. Sharapov1, V. Yavorskij4,5, B. Alper1, R. Barnsley1,P. de Vries1, C. Gowers1, L.-G. Eriksson6, P.J. Lomas1,M.J. Mantsinen7, A. Meigs1, J.-M. Noterdaeme8,9,F.P. Orsitto2 and JET EFDA contributorsa

1 Euratom/UKAEA Fusion Association, Culham Science Centre, Abingdon, Oxon,OX14 3DB, UK2 Associazione Euratom/ENEA/CNR sulla Fusione, Centro Ricerche EnergiaENEA-Frascati 00044, Rome, Italy3 Consorzio RFX—Associazione Euratom-Enea sulla Fusione, I-35127 Padova, Italy4 Euratom/OEAW Association, Institute for Theoretical Physics, University of Innsbruck,Austria5 Institute for Nuclear Research, Kiev, Ukraine6 Association Euratom/CEA, CEA/DSM/DRFC, CEA-Cadarache,13108 St. Paul-lez-Durance, France7 Helsinki University of Technology, Association Euratom-Tekes, PO Box 2200,FIN-02015 HUT, Finland8 Max-Planck IPP-EURATOM Assoziation, Garching, Germany9 Department EESA, Gent University, Belgium

Received 21 January 2005, accepted for publication 5 April 2005Published 28 April 2005Online at stacks.iop.org/NF/45/L21

AbstractGamma-ray images of fast D- and 4He-ions accelerated with third-harmonic ion-cyclotron-resonance heating of4He-beam were simultaneously recorded for the first time in JET tokamak experiments dedicated to the investigationof burning plasmas with 3.5 MeV fusion alpha (α) particles. Gamma (γ ) rays, born as a result of nuclear reactions,9Be(4He, nγ )12C and 12C(D, pγ )13C, between the fast ions and the main plasma impurities, are measured usinga two-dimensional multicollimator spectrometer array, which distinguishes the γ -rays from accelerated D- and4He-ions. Tomographic reconstruction of the γ -ray emission profiles gives images of the fast-ion population in thepoloidal cross-section. The potential of this technique to visualize several energetic ion species and to determinetheir behaviour in different plasma scenarios is demonstrated.

PACS numbers: 52.55.Fa, 52.55.Pi, 52.70.La

(Some figures in this article are in colour only in the electronic version)

The study of fast ions in tokamak plasma experiments isan important part of today’s magnetic fusion programme.It includes development of the ion cyclotron range frequency(ICRF) heating scenarios, which could provide effective

a Annex of Pamela J. et al 2003 Overview of JET results OV-1/1.4, FusionEnergy 2002: Proc. 19th Int. Conf. (Lyon, 2002) (IAEA) Nucl. Fusion43 1540.

mechanisms for additional heating of reactor plasmas, andinvestigation of the behaviour of ions in the megaelectronvoltenergy range for modelling the heating of plasmas byfusion α-particles. Simultaneous measurements of severalfast-ion species are paramount for the burning plasma inITER. At least two types of fast ions are expected in theITER plasma: 1 MeV deuterons from NBI heating andfusion α-particles (4He), which are born in the reaction

0029-5515/05/050021+05$30.00 © 2005 IAEA, Vienna Printed in the UK L21

Letter

D + T → α(3.5 MeV) + n(14 MeV). A principal diagnosticproblem in ITER will be discriminating the NBI deuteriumfrom the fusion α particles. This letter shows for the first timethat a simultaneous measurement of fast D- and 4He-ions ispossible with γ -ray diagnostics.

The γ -ray technique [1–5], now routinely used for thefast-ion study in JET [6, 7], is based on measurements of γ -rays, which are born as a result of nuclear reactions betweenfast ions and the main plasma intrinsic impurities in JET(C, Be). This diagnostic provides information on the energeticion tail and spatial distribution of γ -ray emission.

This letter reports on the first γ -ray images of D- and4He-ions obtained with a two-dimensional array of collimatedγ -ray spectrometers, the so-called gamma cameras. The use ofthe γ -spectrometers has allowed us to essentially increase theefficiency of the measurements, and gave the opportunity todetect γ -rays in the specific adjustable energy windows thatare crucial for distinguishing the emissions from differentfast-ion species in plasmas. A γ -ray image of fast 4He-ionswas recorded during plasma heating with third-harmonic ion-cyclotron-resonance heating (ICRH) of 4He beam-ions [8] inthe experiment, which was dedicated to simulation of burningplasmas with 3.5 MeV fusion α particles. As was shown in[8], a finite Larmor radius effect plays a significant role inthird-harmonic ICRH absorption at the resonance layer. Thewave absorption by resonating ions increases with Larmorradius until a maximum is reached, which typically occursat ion energies in the megaelectronvolt range (ρ ≈ 6 cm).Simultaneously, the D-minority was accelerated by the samethird-harmonic ICRH, and the related γ -ray emission profilewas also measured in a different energy range. Thesemeasurements provide information on the localization ofenergetic D- and 4He-ions in the plasma, and show a cleardifference in γ -ray images depending on the fast-ion orbits ina monotonic and non-monotonic q-profile plasma, providingvery important information for the development of reactorplasma scenarios.

In α-particle simulation experiments with third-harmonicheating of 4He-beams in 4He-plasmas, γ -radiation due to thenuclear reaction 9Be(4He, nγ )12C was detected, showing thesuccessful ICRF-acceleration of 4He-beam ions as intended.The diagnostic capabilities of the reaction 9Be(4He, nγ )12Care determined by the specific reaction cross-section [6]. Thefirst energy level, 4.44 MeV, of the final nucleus, 12C, isexcited by α particles, and as a result the peak at 4.44 MeV(transition 4.44 → 0) in the spectrum appears. The γ -rayemission from the reaction 12C(D, pγ )13C was observed aswell. A peak at 3.09 MeV (transition 3.09 → 0), which isidentified as a γ -emission from the 12C(D, pγ )13C reaction,reflects the presence in the plasma of fast D-ions in themegaelectronvolt range. This evidence indicates that thedeuterium minority also absorbs some ICRF power at the third-harmonic D resonance that coincides with the third-harmonic4He resonance. Figure 1 shows the γ -ray spectra recordedduring 1 s in two similar discharges with a high-resolutionNaI(Tl) spectrometer. It is seen that when the 110 keV neutralbeam heating injector was replaced by the 70 keV one withthe same power, the intensity of the 4.44 MeV γ -emission fellsubstantially, whereas the deuterium γ -peak did not change.This effect can be explained by decreasing single pass 4He

600

400

200

02 4 6 8

Cou

nts

Gamma-ray energy (MeV)

JG04

.528

-1c

12C (D,pγ)13CEγ = 3.1 MeV

9Be (4He,nγ)12CEγ = 4.44 MeV

Figure 1. Gamma-ray spectra measured by the NaI(Tl)detector: solid line—spectrum recorded in discharge with 70 and110 keV 4He-beam injectors; dashed line—spectrum recorded in adischarge with two 70 keV 4He-beam injectors.

2

1

0

3

1.0 2.00 3.0

Rea

ctio

n ra

te (

a.u.

)

Energy (MeV)

JG04

.528

-2bw

12C (D, pγ)13C

9Be (4He, nγ)12C

Figure 2. Calculated reaction-rates for the typical distributionfunctions of D- and 4He-ions deduced from γ -ray measurements inthese experiments.

damping of the ICRF waves due to the smaller value ofthe finite Larmor radius that determines the third-harmonicabsorption. As follows from the γ -ray spectrum analysiswith the GAMMOD code [6], the effective tail temperature,〈THe〉, which is used as a parameter to describe the distributionfunction of fast 4He-ions, falls from 0.5 to 0.35 MeV, whilethe deuterium tail-temperature, 〈TD〉, does not change. Inthese experiments the reaction-rates, R(E) ∝ σ(E)

√Ef (E),

where σ(E) is a nuclear reaction cross-section and f (E)

is a fast-ion energy distribution function, are presented infigure 2. One can see the 4.44 MeV γ -ray emission from

L22

Letter

2.0 2.5 3.0 3.5

0

0.5

1.0

1.0

2.0

0

-1.0

1.5

JG04

.528

-3a

R (m)

Z (

m)

x1011m-3s-1

2.0 2.5 3.0 3.5

0

0.5

1.0

1.5

2.0

1.0

2.0

0

-1.0

JG04

.528

-3b

R (m)

Z (

m)

(×1011m-3 s-1)4He-ions(a) (b) D-ions

Figure 3. Tomographic reconstructions of 4.44 MeV γ -ray emission from the reaction 9Be(4He, nγ )12C (a) and 3.09 MeV γ -ray emissionfrom the reaction 12C(D, pγ )13C (b) deduced from simultaneously measured profiles.

2.6

-0.2

0

0.2

0.4

0.6

0.8(a) (b)

2.8 3.0 3.2 3.4 3.6R (m)

Z (

m)

JG04

.528

-4a

1.9 MeV

2.6

-0.2

0

0.2

0.4

0.6

0.8

2.8 3.0 3.2 3.4 3.6R (m)

Z (

m)

JG04

.528

-4b1.5 Mev

2.0 Mev

1.0 Mev

Figure 4. Calculated orbits of 4He-ions with energy 1.9 MeV (a) and D-ions with energy 1, 1.5 and 2 MeV (b) in the same magneticconfiguration as in the experiment and with v‖ close to zero at the resonance layer, Rres = 3 m.

the 9Be(4He, nγ )12C reaction is produced by 4He-ions in thenarrow energy band around an energy of 2 MeV. At the sametime, a rather broad band of D-ions (0.8–2.5 MeV) gives rise tothe 3.09 MeV gamma-rays due to the reaction 12C(D, pγ )13C.

Spatial profiles of the γ -ray emission in the energyrange Eγ > 1 MeV have been measured in JET using thegamma cameras, which have ten horizontal and nine verticalcollimated lines of sight. Each collimator defines a poloidal-viewing extent at the centre of the plasma of about 10 cm. The

detector array comprises 19 CsI(Tl) photo-diodes (10 mm ×10 mm × 15 mm). The CsI(Tl) detectors are well calibratedwith radioactive sources, 22Na (511, 1275 keV), which areembedded in the detector-array module. The data acquisitionsystem accommodates the γ -ray count-rate measurement infour independently adjustable energy windows. This allowsallocating γ -ray peaks for a given fast-ion population inspecific windows to count them separately. The most recentMonte Carlo calculations of the detector response functionshows that the size of the CsI(Tl) scintillator is suitable for

L23

Letter

2.0 2.5 3.0 3.5

0

1

2

3

4

5

1.0

2.0(a) (b)

0.0

-1.0

JG04

.528

-5a

R (m)

Z (

m)

(×1011 m-3 s-1)

2.0 2.5 3.0 3.5

0

2

4

6

8

10

1.0

2.0

0

-1.0

JG04

.528

-5b

R (m)

Z (

m)

Figure 5. Tomographic reconstructions of 4.44 MeV γ -ray emission from the reaction 9Be(4He, nγ )12C related to profiles measured during1 s in the non-monotonic (a) and monotonic (b) q-profile phases of the same plasma discharge.

γ -ray measurements in the megaelectronvolt range. It isfound that a noticeable contribution to the γ -ray spectrum withenergy Eγ � 2mec

2 gives so-called single- and double-escapepeaks, which result from the high probability of e−e+-pairgeneration in the detector. The double-escape peak efficiencyof this detector is about 3%, which in this case is sufficientfor reliable counting. The experiments were performed in a4He-plasma with deuterium as a minority. This is a favourablebackground conditions case, which we use for precise γ -raymeasurements. Neutron rate monitoring with the NE213detector-array installed in the cameras shows that the neutroncount-rate does not exceed 1 kHz.

In the gamma cameras the measurable line-integral γ -raybrightness along the viewing direction can be represented in asimplified form as

� ∝∫

E

σ(E)√

E dE

∫L

nZ(r)F (E; r, θ) dL, (1)

where F(E; r, θ) is a fast-ion distribution function, dependingon the ion energy, minor radius and poloidal angle and nZ(r)

is a low-Z impurity density. Experimental data obtained forthe 19 lines of sight are tomographically reconstructed to getthe local gamma-ray emissivity in a poloidal cross-section,I (r, θ) ∝ nZ

∫E

F(E; r, θ)σ (E)√

E dE. In these reconstruc-tions the full geometry of the collimators and detectorefficiencies were taken into account, but small effects ofattenuation and scattering of the γ -rays were neglected. It isassumed that the distribution of the low-Z impurities is uniformin the plasma core as confirmed by atomic spectroscopymeasurements. For the tomographic reconstruction aconstrained optimization method [9] is used, which was

successfully applied earlier to soft x-ray, bolometer and γ -raymeasurements at JET. Reliability assessments have been doneby varying the reconstruction parameters within a reasonablerange as described in [6, 9]. The effective spatial resolution ofthe diagnostic in these experiments is about ±6 cm.

The special energy windows, containing the 3.09 and4.44 MeV peaks with their single- and double-escape satelliteswere set up to measure the spatial profiles of the γ -rayemission from D- and 4He-ions. A typical example of thetomographic reconstruction of the measured line-integratedprofiles recorded during 1 s in a steady state phase of the plasmadischarge is shown in figure 3. It is seen clearly that the γ -rayemission profile produced by fast D-ions (figure 3(b)) differsfrom the profile from 4He-ions (figure 3(a)). This effect can beexplained by the difference in pitch-angle distribution between4He beam-ions injected into the plasma quasi-tangentially andisotropic D-minority ions. Furthermore, as seen in figure 2,the D-minority ions, unlike the 4He-ions, produce γ -rays ina broad band of energies, and so the D-ions with differentorbit topologies contribute to the γ -ray emission profile.Figure 4 shows the typical orbits of the ICRF acceleratedD- and 4He-ions (with v‖ close to zero at the resonance layer,Rres = 3 m), which mainly contribute to the observed γ -rayemission. They were calculated for the magnetic configurationfor this particular discharge. For the D-ions the orbits werecalculated in a broad energy range: 1.0, 1.5 and 2.0 MeV. Forthe 4He-ions the most probable energy value, 1.9 MeV, waschosen in accordance with figure 2.

The ability to separate γ -rays from different energy bandsallows a study of the fast-ion behaviour in some importantplasma scenarios with the gamma cameras. For instance,

L24

Letter

Figure 6. Calculated orbits of the 4He-ion in discharge with anon-monotonic q-profile.

a unique γ -ray emission profile produced by 4He-ionswas measured in an advanced regime plasma discharge.In this discharge, Alfven cascade and reverse-shear sawtoothinstabilities were observed, indicating a non-monotonic safetyfactor (q) profile [10]. According to theoretical analysisof fast α-particle orbit topology in non-monotonic q-profileplasmas [11], very broad confined banana orbits are expectedfor the 4He-ions with energies around 2 MeV. The γ -rayemission profile presented in figure 5(a) is consistent withthis prediction. Actually, the calculated orbits (typically,with v‖ close to zero at the resonance layer, Rres = 3 m,and an almost stagnation orbit) of a probe 1.9 MeV 4He-ion in a plasma configuration with deeply reversed magneticshear (figure 6) demonstrates that the model with a moderatecurrent hole size (r < 0.3a) corresponds to the measured γ -ray emission profile. It is expected that the 4He-ions withnear-stagnation orbits mostly contribute to the γ -ray emissionobserved in the region R > Rres. In the same discharge,in the phase with a monotonic q-profile, a γ -ray emissionprofile was observed (figure 5(b)), which is consistent with thecalculated standard banana-orbit as presented in figure 4(a).

It is worth stressing that in a long-term perspectivesimilar measurements with two-dimensional cameras could beused in burning-like ITER plasmas, but the γ -ray detectorsshould be protected against severe neutron emission withspecial neutron filters [12]. Since the main impurity in ITERplasmas will probably be beryllium, the slowed deuteronswill give rise to gamma rays (2.88 and 3.37 MeV) fromthe 9Be(D, nγ )10B and 9Be(D, pγ )10Be reactions, whereasthe fusion alpha particles with energies around the 2 MeVresonance in the 9Be(4He, nγ )12C reaction produce 4.44 MeVgamma rays. Using γ -ray spectrometers in every channel ofthe ITER cameras, the α-particle slowing down profile can bemeasured with the technique successfully tested in JET tracetritium experiments [13]. Simultaneous measurements of theNBI power deposition and α-particle slowing down profiles arevery important for optimization of different plasma scenariosand understanding the fast-ion confinement effects for furtherprogress on the way to DT-ignition.

This work has been conducted under the European FusionDevelopment Agreement and is partly funded by Euratomand the United Kingdom Engineering and Physical SciencesResearch Council. The authors wish to acknowledge fruitfuldiscussions with Dr A.W. Morris.

References

[1] Medley S.S. and Hendel H. 1982 Bull. Am. Phys. Soc.26 980

[2] Newman D.E. and Cecil F.E. 1984 Nucl. Instrum. Methods227 339

[3] Cecil F.E. and Medley S.S. 1988 Nucl. Instrum. MethodsPhys. Res. A 271 628

[4] Sadler G.J. et al 1990 Fusion Technol. 18 556[5] Jarvis O.N. et al 1996 Nucl. Fusion 36 1513[6] Kiptily V.G. et al 2002 Nucl. Fusion 42 999[7] Kiptily V.G. et al 2003 Rev. Sci. Instrum. 74 1753[8] Mantsinen M. et al 2002 Phys. Rev. Lett. 88 105002[9] Ingesson L.C. et al 1998 Nucl. Fusion 38 1675

[10] Sharapov S.E. et al 2002 Phys. Plasmas 9 2027[11] Yavorskij V. et al 2003 Nucl. Fusion 43 1077[12] Kiptily V.G. et al 1998 Tech. Phys. 43 471[13] Kiptily V.G. et al 2004 Phys. Rev. Lett. 93 115001

L25