Fusion Engineering and Design 81 (2006) 163–168

Applicability of �-ray-induced X-ray spectrometry to in situmeasurements of tritium retention in plasma-facing

materials in ITER

M. Matsuyamaa,∗, Y. Torikaia, N. Bekrisb, M. Gluglab, A. Erbec,W. Naegelec, N. Nodad, V. Philippse, P. Coadf, K. Watanabea

a Hydrogen Isotope Research Center, Toyama University, Gofuku 3190, Toyama 930-8555, Japanb Tritium Laboratory, FZK, EURATOM Association, D-76021 Karlsruhe, Germany

c Intitute for Materials Research II, FZK, EURATOM Association, D-76021 Karlsruhe, Germanyd National Institute for Fusion Science, Oroshi-cho, Toki-shi, Gifu 509-5292, Japan

e Institute for Plasma Physics, FZJ, EURATOM Association, TEC, 52425 Julich, Germanyf EURATOM/UKAEA Fusion Association, Culham Science Center, Abingdon, Oxon OX14 3DB, UK

Received 31 January 2005; received in revised form 30 August 2005; accepted 30 August 2005Available online 20 December 2005

Abstract

facingsione

ioactivityupportsserved onretention

untlkesin

Applicability of �-ray-induced X-ray spectrometry (BIXS) for in situ measurements of tritium retention by plasma-materials in ITER was examined using two�-emitters and a divertor tile with metallic supports employed during D–T fuexperiments in JET. Measurements of the tile with a�-emitter showed that the presence of the�-field has no influence in the shapof the�-ray-induced X-ray spectra, although the whole intensity in the observed energy region was dependent on radof the�-emitters. To simulate more accurately the in-vessel environment of PFMs, one of the divertor tiles with metallic swas subjected to BIXS measurements. Although the metallic supports were gamma active, no significant effect was obthe recorded X-ray spectrum. These results clearly indicate that BIXS can be applied to in situ measurements of tritiumby PFMs with a simple shielding of the X-ray detector.© 2005 Elsevier B.V. All rights reserved.

Keywords: Tritium retention; In situ measurement; BIXS; Plasma-facing materials; ITER

∗ Corresponding author. Tel.: +81 76 445 6926;fax: +81 76 445 6931.

E-mail address: masao@hrc.toyama-u.ac.jp (M. Matsuyama).

1. Introduction

It is of a great importance to assess the amoof tritium retained by the surface and/or in the buof plasma-facing materials (PFMs) of fusion devicfrom viewpoints of control of fuel particle balance

0920-3796/$ – see front matter © 2005 Elsevier B.V. All rights reserved.doi:10.1016/j.fusengdes.2005.08.037

164 M. Matsuyama et al. / Fusion Engineering and Design 81 (2006) 163–168

the reactor core as well as safety and economy of tri-tium. For this purpose, the applicability of the newlydeveloped technique of�-ray-induced X-ray spectrom-etry (BIXS) has been previously examined[1–4]. Thistechnique is based on the measurement and analysisof X-ray spectra induced by the�-rays from tritiumtrapped in/on materials, and main features are as fol-lows: non-destructive measurements, discrimination ofsurface layers and bulk, and depth profile analyses oftritium are possible. Feasibility of BIXS was exam-ined using graphite samples irradiated with a givenamount of tritium ions[2], and it was found that BIXScan measure the tritium amounts retained on the sur-face layers and in bulk of graphite independently. Inaddition to this, it was shown that the depth profileof tritium in the bulk could be clarified by computersimulation.

Application of BIXS to non-destructive measure-ment of the tritium amounts was further extended basedon the above results. The amount and distribution of tri-tium retained on/in the carbon fiber composite (CFC)tiles exposed to tritium plasmas in the JET machine dur-ing the DTE1 campaign were measured in detail[3,4].It was concluded from this examination that the presenttechnique is significantly useful for in situ evaluation oftritium retention on/in PFMs as well as metallic impu-rities deposited on PFMs.

To apply BIXS for in situ measurements of tritiumretained by PFMs in ITER, effects of�- and X-raysemitted from the neutron-activated structure materialso nce.S el sitya t thea suei ed X-rp ectrai it-t tedm

2

inet iceo ter-

actions between�-rays of tritium and materials: (1)measurements of radiation spectra in an energy regionbelow 18.6 keV by using two�-emitters (137Cs and60Co) as models of activated materials, (2) prelim-inarily tests using both the cylindrical carbon fibercomposite (CFC) sample contained a given amount oftritium and137Cs, and (3) measurements of X-ray spec-tra emitted from a CFC tile with the activated metallicparts, which were exposed to neutrons produced byD–T fusion experiments in the JET machine. All themeasurements of CFC tiles were performed in a glovebox of the hot cells in the Forschungszentrum Karl-sruhe.

A semiconductor detector equipped with a highpurity Ge crystal was used for measurements of radia-tion spectra such as�-rays and�-ray-induced X-rays.Energy resolution of the present X-ray detector wasdetermined to be 129 eV at 5.9 keV by using a55Fesource. The radiation entrance window of the X-raydetector was made of a specially designed thin beryl-lium plate (8�m in thickness) to gain an effectivetransmittance of low energy X-rays. The size of theGe crystal employed for the present detector was 8 mmin diameter and 5 mm in thickness.

The activities of137Cs and60Co sources used were2.2 MBq and 1.5 TBq, respectively. In the first measure-ment, since activity of the former�-emitter was not sohigh, it was fixed near the X-ray detector surroundedwith a few lead bricks in the glove box and the radiationspectrum from the137Cs source was measured. How-e dn rgea edie tionw iont X-r ll ass torw fors

pleu waso theD plew ctraw dt

n the BIXS spectra must be elucidated in advaince X-rays induced by tritium�-rays appear in th

ow energy region below 18.6 keV and have intennd spectral shapes that give information aboumount and depth profile of tritium, an important is

s to examine the changes in a shape of the observay spectra with the coexistence of X- and�-rays. In theresent study, therefore, changes in the X-ray sp

nduced by tritium�-rays were examined using emers of�-rays and a CFC tile attached with the activaetallic supports.

. Experimental

Three kinds of tests were carried out to examhe effects of ambient radiations in a fusion devn measurements of X-ray spectra induced by in

ver, radiation spectrum from the60Co source coulot be measured by a similar way owing to the lactivity. That is, the60Co source was usually stor

n a thick lead-cell. Thus, radiation from the latter�-mitter was measured by opening slightly a protecindow of a specially designed lead-cell. In addit

o a slight opening of the protection window, theay detector was placed at outside of the lead-cehown inFig. 1. The rear side of the X-ray detecas also surrounded with a number of lead bricksafety.

In the second series of tests, a cylindrical samsed was cored out from the 1IN1 CFC tile thatne of the divertor tiles exposed to D–T plasmas inTE1 campaign of JET. Size of the cylindrical samas 6 cm in diameter and 3 cm in length. X-ray speere measured by placing the137Cs source just behin

he cylindrical sample.

M. Matsuyama et al. / Fusion Engineering and Design 81 (2006) 163–168 165

Fig. 1. Photograph of the location of the X-ray detector for measure-ments of the60Co source.

In the third series of tests, the 1BW5 CFC diver-tor tile, which was located at the dome top of theMkIIA configuration in the JET machine, was used formeasurements. During the measurement of the X-rayspectrum, the distance between the beryllium windowof the X-ray detector and the sample surface was keptat 5 mm, and argon gas was flowed in this narrow gapin order to improve the detection efficiency of tritiumretained in surface layers (average escape depth of�-rays: about 0.5�m) of the divertor tile. A typicalconfiguration of a CFC sample and X-ray detector inthe glove box is shown inFig. 2.

Fig. 2. Photograph of a typical configuration of a CFC tile and X-rayd

Fig. 3. Radiation spectra from a137Cs source: (A) is the whole rangeand (B) is the region of interest.

3. Results and discussion

3.1. Measurements of radiation spectra of 137Csand 60Co

Figs. 3 and 4show the radiation spectra of137Csand60Co measured in this study. Multiple peaks anda continuous spectrum were observed in both spectra.The peaks were attributed to K� and K� X-rays of Pbatoms from energy analysis, which were emitted byphotoelectric effects of�-rays. The continuous spec-trum consisted of radiation corresponding to Comptonscattering of�-rays from137Cs and60Co, in additionto natural radiation. As shown in (B) of both figures,the continuous spectra exhibited a simple shape in theregion of interest below 18.6 keV, although the intensityand energy of the radiation of both�-emitters was quitedifferent. A similar simple increase in the region ofinterest will be seen even in the ITER case, though theintensity of a continuous spectrum may be largely dif-

etector in the glove box.

166 M. Matsuyama et al. / Fusion Engineering and Design 81 (2006) 163–168

Fig. 4. Radiation spectra from a60Co source: (A) is the whole rangeand (B) is the region of interest.

ferent from the present observation. It was suggested,therefore, that it is possible to subtract the effects ofCompton scattering from the observed X-ray spectruminduced by�-rays of tritium.

3.2. Effects of γ-rays from 137Cs on a shape ofβ-ray-induced X-ray spectrum

To confirm the feasibility of the subtraction methoddescribed above,�-ray-induced X-ray spectra weremeasured using a cylindrical sample of the 1IN1 tile asa preliminarily test, where the137Cs source was placedjust behind the cylindrical sample. The observed spec-trum is shown inFig. 5and is made up from not onlythe contributions of the tritium species but also five dif-ferent metallic species deposited on the surface layersof the cylindrical sample. In this spectrum, contribu-tion of the Compton scattering of�-rays was estimatedby the dotted line, which was drawn by based on theresult ofFig. 3. In the present conditions, although the

Fig. 5. Radiation spectrum observed for the cylindrical tile samplewith a 137Cs source.

effects of the Compton scattering were not so largein comparison with a whole intensity of the�-ray-induced X-rays, the spectrum with the intensity of thedotted line was subtracted was in good agreement witha spectrum obtained from the cylindrical sample only.This indicates that the presence of the�-field has noinfluence in the shape of the�-ray-induced X-ray spec-tra, although a whole intensity in the observed energyregion is dependent on radioactivity of the�-emitters.It is suggested, therefore, that it is possible to evaluatequantitatively the amount and depth profile of tritiumfrom the observed spectra.

3.3. Effects of radioactive metallic supportsattached with a CFC tile on BIXS

To simulate more accurately the in-vessel environ-ment of PFMs, one of the divertor tiles (1BW5 tile) withmetallic parts was subjected to BIXS measurements.Photographs of the tile surfaces are shown inFig. 6.Numbers in the photograph (Fig. 6A) indicate locationof measuring spots.Fig. 7 shows an example of theobserved radiation spectra in a region above the maxi-mum energy of�-rays of tritium, which was measuredat the plasma-facing surface of the tile. The observedradiation is basically due to the activated metallic sup-port of the CFC tile. The radiation consisted of�-rays and its Compton scattering. The radiation peaksappeared at 122.4 and 136.8 keV, which correspond to� 57 gi dia-t

-rays emitted from a nucleus ofCo atoms. Takinnto account the intensity of Compton scattering raion, it is expected that there are some peaks of�-rays

M. Matsuyama et al. / Fusion Engineering and Design 81 (2006) 163–168 167

Fig. 6. Photographs of the CFC tile (1IBW5): (A) is the plasma-facing surface and (B) is the rear surface. Numbers in the upperphotograph show the measuring spots by BIXS.

in a higher energy region than 160 keV, which cannotbe detected by the present X-ray detector.

Fig. 8shows the spectra observed for three spots onthe plasma-facing surface. Characteristic X-ray peaksof Ar K�, Cr K�, Fe K� and Ni K� and bremsstrahlungX-rays can be clearly seen in the observed spectra. Theamount of tritium retained in surface layers of the tileand the depth profiles of tritium in bulk were estimatedfrom the observed spectra, assuming any impedimentfrom�-rays and their Compton scattering radiation wasnot so serious for evaluation of tritium. That is, therewere not any significant effects for the recorded X-ray

Fig. 7. Radiation spectrum observed for the 1BW5 tile.

Fig. 8. X-ray spectra observed for three spots on the plasma-facingsurface of the 1BW5 tile.

spectra, although the metallic supports were gammaactive. These results clearly indicate that BIXS is appli-cable for in situ measurements of tritium retained byPFMs with a simple shielding of the X-ray detector,even in presence of a strong�-field, such as is expectedin ITER.

3.4. Design of an X-ray detector for application ofBIXS to ITER

The Ge crystal used in the present study to detectX-rays induced by�-rays was 5 mm in thickness. How-ever, to detect X-rays of lower energy than 18.6 keV,which is the maximum energy of�-particles from tri-tium, the present crystal is too thick. If the aim of theX-ray detector is merely to detect radiations of lowerenergy than 18.6 keV, 0.2 mm in thickness is enoughfor efficient detection of X-rays mentioned above. Theabsorption fraction of the X-rays in the Ge crystal ishigher than 99%, based on a mass absorption coeffi-cient of photons for Ge (the density is 5.36 g/cm3 at293 K) of 51.1 cm2/g at 18.6 keV. Namely, the reduc-

168 M. Matsuyama et al. / Fusion Engineering and Design 81 (2006) 163–168

tion in thickness of a Ge crystal is able to decreaseeffects of�-field on BIXS measurements, because across-section of the high-energy photons with Ge willexponentially decrease with decreasing the thicknessof a Ge crystal. In addition to the improvement of aGe crystal, the effects of radiations from the neutron-activated structural materials can be lowered below10% by shielding with a given thickness of the leadmaterials in comparison with no shielding, althoughthe precise reduction rate depends on the energy ofphotons.

4. Summary

From the viewpoints of control of fuel particle bal-ance in the reactor core as well as safety and economy oftritium in the fusion devices, it is of a great importanceto assess the tritium retention by the surface and/or inthe bulk of plasma-facing materials (PFMs). To apply�-ray-induced X-ray spectrometry (BIXS) for in situmeasurements of tritium retention by PFMs in ITER,possible effects of�- and X-rays emitted from theneutron-activated structure materials on the BIXS mea-surements have been examined using�-emitters and acarbon fiber composite tile with radioactive metallicsupports.

In the first test, two�-emitters (137Cs and60Co) wereused as a model, and the spectra in the energy regionbelow 18.6 keV were measured using a high purity Ge

detector. The observed spectra consisted of radiationdue to Compton scattering of�-rays in addition to thenatural radiation. Measurements with two�-emittersshowed that the presence of the�-field has no influencein the shape of the�-ray-induced X-ray spectra.

To simulate more accurately the in-vessel environ-ment of PFMs, one of the divertor tiles with metal-lic supports was subjected to BIXS measurements.Although the metallic supports were gamma active,there is not any significant effect observed for therecorded X-ray spectrum. It was suggested from theseresults that BIXS could be applied to in situ mea-surements of tritium retention by PFMs with a simpleshielding of the X-ray detector, even in the presence ofa strong�-field as in ITER.

References

[1] M. Matsuyama, K. Watanabe, K. Hasegawa, Tritium assayin materials by bremsstrahlung counting method, Fusion Eng.Design 39/40 (1998) 929–936.

[2] M. Matsuyama, T. Murai, K. Watanabe, Quantitative measure-ments of surface tritium by�-ray-induced X-ray spectrometry(BIXS), Fusion Sci. Technol. 41 (2002) 505–509.

[3] M. Matsuyama, N. Bekris, M. Glugla, N. Noda, V. Philipps,K. Watanabe, Non-destructive tritium measurements of MkIIAdivertor tile by BIXS, J. Nucl. Mater. 313–316 (2003) 491–495.

[4] Y. Torikai, M. Matsuyama, N. Bekris, M. Glugla, P. Coad, W.Naegele, et al., Tritium distribution in JET Mark IIA type divertortiles analysed by BIXS. J. Nucl. Mater 337–339 (2005) 575-579.