Tritium retention in plasma-facing components 9 - Shimada - Tritium...Masashi Shimada Fusion Safety...

ww

w.in

l.gov

Masashi Shimada

Fusion Safety Program Idaho National Laboratory

Tritium retention in plasma-facing components

38th U.S. Tritium Focus Group Meeting at PNNL

May 9, 2017 STI: INL/MIS-17-41957

Outline of the talk • Motivation and challenge

• Methodology and research approach

• Progress and future plans

M. Shimada (INL) | 38th TFGM | May 9, 2017 | PNNL 2 Joint European Torus (JET)

ITER first wall/divertor

M. Shimada (INL) | 38th TFGM | May 9, 2017 | PNNL 3

Beryllium first wall (low Z è low radiation loss)

Tungsten divertor (high melting point, low sputtering yield and high thermal conductivity)

Pbremsstrahlung ~ (Zeff )2

•  Plasma-facing component (PFC) –  Beryllium at first wall to minimize Zeff –  Tungsten at divertor for high qheat>10 MW/m2

Material challenges in plasma-facing components

4

as well as to account for the processes of excitation, ionization, andcharge-exchange that can result in species re-deposition. Likewise,the interface between the surface and the bulk, where defectcreation is no longer influenced by the presence of a free surface,is critical in determining the extent to which defect creation byhigh-energy neutrons impact retention and permeation of hydro-gen isotopes, with a significant unknown existing with respect tothe tritium permeation behavior in metallic PFC at elevatedtemperatures.

Gaining a physical understanding and establishing a predictivemodeling capability in this critical PSI area requires that complexand diverse physics occurring over a wide range of length(Ångströms to meters) and time (femtoseconds to seconds, daysto years) scales be addressed simultaneously, and that extensivephysical processes across the plasma–surface-bulk materials inter-faces be integrated. Figs. 1 and 2 illustrate phenomena that governthe response of the materials surface to plasma exposure [9], andthe computational models that must be accurately integrated.While vastly different length scales characterize the surface(!nm) and plasma processes (!mm) as indicated in Fig. 1, theplasma and the material’s surface are strongly coupled to eachother, mediated by an electrostatic and magnetic sheath, throughthe nearly continuous exchange and recycling of incident ion andneutral species and the re-deposition of eroded particles. Theseinteractions are more explicitly shown in Fig. 2, along with the cor-responding time scales upon which they occur. These physical pro-cesses occur over a disparate range of time scales, which poses achallenge both to modeling, and experimental characterization ofboth the individual and coupled processes. As one example, the highprobability (>90%) of prompt local ionization and re-deposition ofsputtered material atoms means that the surface material that isin contact with the plasma is itself a plasma-deposited surface, asopposed to the original well-ordered surface of the material thatexisted at the beginning of operation [9]. Likewise, the recyclingof hydrogen plasma (fuel) is self-regulated through processesinvolving near-surface diffusion, trapping, and gas bubble forma-tion, coupled to the ionization that results from interactions withthe plasma. The multitude of time and length scales controllingmaterial evolution and device performance requires the develop-ment not only of detailed physics models and computational

strategies at each of these scales, but also of computational algo-rithms and methods to couple them strongly and in such a way thatcan be robustly and vigorously tested and validated. It is importantin this regard that PFC simulation tools capture the kinetic evolutionof defect and impurity species over diffusional timescales that areinaccessible through molecular dynamics (MD) techniques alone.

As helium, deuterium or tritium particles bombard the surface,they can reflect, induce sputtering of surface atoms, be adsorbedonto the surface, or implanted below the surface depending onthe type of ion, and their kinetic energy and angle of incidence.Likewise, sputtered or eroded material from a surface can be ion-ized, transported through the plasma and re-deposited. Sinceimplantation energies are generally in the range of 10–1000 eV,the implantation depth is generally only a few nanometers. Asmore implanted particles accumulate within the surface layer,eventually a steady-state condition can result, in which the fluxof species implanted into the materials is balanced by that releasedfrom the material. The extent to which both surface morphologyand sub-surface defect creation and evolution processes drivenby neutron-induced damage influence the diffusion, trapping andprecipitation of hydrogen and helium species into gas bubbles isan outstanding question that impacts the tritium permeation,retention and near-surface saturation levels.

Tungsten has recently been selected as the sole divertor mate-rial in ITER [10,11], and is the leading candidate material for DEMOand future fusion reactors. Laboratory experiments performed inlinear plasma devices indicate the possibility of substantial surfacemodification in tungsten exposed to low-energy, helium plasma, ormixed helium–hydrogen plasma, although the observed surfaceresponse is strongly temperature-dependent and likely dependenton the ion energy and flux. Pitted surfaces are observed below"1000 K [12], whereas a ‘‘nanostructured,’’ low-density ‘‘fuzz’’ or‘‘coral’’ surface morphology is observed between approximately1000 and 2000 K [13–16], while micron-sized holes, or pits, areobserved to form above about 2000 K [17,18]. The nanostructured‘‘fuzz’’ has recently been observed in the divertor regions of a toka-mak device operating with a helium plasma as well [19]. Suchsurface features could lead to changes in heat transfer, fuel(deuterium/tritium) retention [20], increased rates of erosionthrough both sputtering and dust formation [21], and

Fig. 1. Schematic illustration of the synergistic plasma surface interaction processes that dictate material evolution and performance in the magnetic fusion plasmaenvironment, as reproduced from [9].

B.D. Wirth et al. / Journal of Nuclear Materials 463 (2015) 30–38 31

•  Materials under extreme condition: –  Intense plasma particle/heat flux –  14 MeV neutron flux

•  Safety concern for tritium inventory: –  4 kg: ITER site limit (~ 1,500,000 TBq)*

•  2~3 kg: in tritium plant ? •  1 kg: in hot cell •  1 kg: in vacuum vessel including piping system

–  (120 -150) g in cryo-pumps –  (850 – 880) g in plasma-facing components

–  ITER: Be co-deposit is predominating factor. –  DEMO: Neutron-irradiated W will be predominant factor.

structure (A) and ring-like structure (B), see Fig. 1c). In these pro-cesses, growth of the highly pressurized He bubbles in the surfaceregion could lead to formation of protrusions, the size of which de-pends on the size of bubbles. Moreover, blister-like structures canform in some parts by the development of the He bubble growth.It is thought that active surface diffusion of tungsten atoms playsa key role in forming protrusions and fine structures, becausesurface diffusion is, in general, much faster than lattice diffusion.It is likely that the surface diffusion allows W to combine with Hebubbles to change the surface morphology.

The nature of the growth of nano-structured layers with the Heflux of (4–6) ! 1022 m"2 s"1 at 1120 K in the PISCES linear plasmadevice is revealed by the scanning electron microscope (SEM)images in Fig. 2 [5]. As the exposure time is increased, the nano-structured layer grows in thickness to more than 5 lm for the timeof 2.2 ! 104 s. Layer thickness data are shown for exposure tem-perature regimes of 1120 and 1320 K. For both exposure tempera-tures, the data are well characterized by a square root dependenceon time, suggesting the diffusion-like process dominates nano-structure growth. Similar dependence was also seen in the NAGDISlinear plasma device with an incubation He fluence of about4 ! 1024 m"2 [14]. In addition, formation of the nano-structure ishardly affected by simultaneous D implantation [15]. Althoughthe growth mechanism of the nano-structure is not well under-stood, a model called ‘‘Visco-elastic model’’, which predicts thet1/2 growth of the nano-structure, was recently proposed by Kra-sheninnikov [16].

Formation conditions for the nano-structure were reported byKajita et al. [13]. The incident ion energy and surface temperatureare key parameters for the formation. Fig. 3 shows the formationwindow of the nano-structure. The gray colored region and thehatched region correspond to the conditions in which nano-struc-ture and large helium bubbles, respectively, are observed after he-lium irradiation. The formation window ranges between 1000 Kand 2000 K, and over the ion energy of about 20 eV. The thicknessof the nano-structure tends to increase with temperature. The en-ergy threshold is #12 eV for the nano-structure formation on VPS(vacuum vapor deposition) W on graphite [4]. This threshold en-ergy is somewhat higher than the surface barrier potential of Wfor He (#6 eV) [3]. The reason why the energy threshold for thenano-structure formation (12–20 eV) is higher than the barrier po-tential is not known. The formation rate of the nano-structure maybe too low for detection if the energy lies below the thresholdenergies.

In ITER, heat and particle flux at the outer strike point would beabout 10 MW m"2 and 1024 m"2 s"1 [17]. In NAGDIS and PISCES,the highest He flux is about 1023 m"2 s"1, which would be aboutthe same as the highest He flux to divertor plates in ITER, becauseHe concentration in burning plasmas would be 5–10%. The heatflux to W specimens in these linear plasma devices, however, isabout an order of magnitude lower than to ITER. He irradiationwith comparable heat and particle flux to ITER divertor has been

Fig. 1. He irradiated W in NAGDIS-II at 1400 K and 50 eV. He fluences are (a), (b)1.1 ! 1025 m"2 and (c) 2.4 ! 1025 m"2 [14].

Fig. 2. Cross-sectional SEM images of W targets exposed to pure He plasma at 1120 K for exposure times from (a) 300 s to (e) 22,000 s with the He flux of (4–6) ! 1022

m"2 s"1 [5].

S268 Y. Ueda et al. / Journal of Nuclear Materials 442 (2013) S267–S272

samples were irradiated at approximately 500 !C, the precipitatesize in the samples irradiated at approximately 800 !C was largerthan those in the samples irradiated at 500 !C. This differencemight be attributed to the growth of defect clusters at highertemperatures by the higher diffusivity of solute elements.

3.3. Damage structure map of irradiated W

On the basis of our irradiation data and the previously publisheddata obtained by EBR-II irradiation [3e6], damage structure mapsof W and WeRe alloys are suggested, as shown in Fig. 2. Voids arethe major damage structure of pure W for irradiation at tempera-tures greater than 400 !C and at damage levels above 0.17 dpa.Dislocation loops are also observed in lower dpa region, but theirnumber density is less than that of voids. The dashed lines areboundary lines of predicted microstructural regions. For example,the right-hand side of a solid line for a void lattice shows the void-lattice-observed region. The right-hand sides of “5/10Re” dashedline and “26Re” chain line also show the precipitate-observed re-gions of W-5%Re and W-26%Re alloys, respectively. Our datashowed that void lattices were observed in pure W irradiated atlevels as high as 0.96 dpa at 538 !C and at levels as high as 1.54 dpaat 750 !C [13]. The formation mechanism of the void lattice has notyet been clarified [22] but the lattice structure is known to beformed under strict balances of vacancy and interstitial productionand annihilation. In the present work, we considered that as long asthe production and annihilation of point defects in the structure arebalanced and the lattice structure is stable. With increasing irra-diation fluence, transmuted Re increases and it enhances vacancy-

interstitial recombination by Re effects, and then the balance ofproduction and annihilation of defects changes. Finally, the voidlattice structure change. The structure change after the higherirradiation level will be discussed later.

In the irradiated WeRe alloys, acicular precipitates were themajor structure and the precipitation behavior depended on the Recontent and on the damage level. Precipitations in W-26%Re wereobserved at damage levels greater than 0.2 dpa. In W-5/10%Re,precipitates were observed at damage levels greater than 0.5 dpa.In W-3%Re, microstructure data of irradiation sample was notenough to determine the precipitate line.

3.4. Irradiation hardening

The relation between irradiation hardening and the Re contentof starting materials after JMTR, Joyo and HFIR irradiation areshown in Fig. 5 [13]. The irradiation temperature region is400e600 !C for Fig. 5(a) and it is 740e800 !C for Fig. 5(b). Theresults show that at lower dpa levels (lower than 0.5 dpa), theirradiation-induced hardening of Wwas relatively higher than thatof WeRe alloys. The hardening of W was caused by the void anddislocation loop formations, as shown in Fig. 1. The lower irradia-tion hardening of WeRe alloys compared with pure W in this dparange might be attributable to suppression of the formation ofvoids and dislocation loops in WeRe alloys.

The hardening of the WeRe alloys increased after approxi-mately 1 dpa irradiation, as shown in Fig. 5(a) and (b). The results ofmicrostructural observation show that fine and dense precipitatesformed in the hardened specimens and may work as obstacles to

Fig. 4. TEM micrographs of neutron irradiated pure W and W-10%Re alloy at 750 !C in the Joyo reactor ((a) and (b)) and at 800 !C in the HFIR ((c) and (d)).

A. Hasegawa et al. / Journal of Nuclear Materials 471 (2016) 175e183 179

1.�BackgroundWe need to control tritium (T) inventory inneutronͲirradiated first walls (FWs), which mayrelease T through the generation of decay heatunder the conditions of lossͲofͲcoolantaccidents of fusion reactors.

Tungsten (W) materials are primary candidatesof FWs.

T retention in W is dominated by trapping effects in radiation defects such asvacancies and voids, because the solubility of hydrogen isotopes in W lattice isextremely low.

To date, the trapping effects by radiationdefects have been mainly examined usingion irradiation techniques.

SlimCS by JAEA

Vacancy

VacancyͲcluster�(voids) T

Data for nͲirradiated W has started to begained in JapanͲUS joint research projectTITAN.

2

* Fushikuma Daiichi released ~ 540,000 TBq in 2011

Outline of the talk • Motivation and challenges




High Flux Isotope Reactor (HFIR), ORNL

Research approach for tritium retention study •  Challenges:

–  Unavailability of high-flux 14 MeV neutron source. –  Unavailability of simultaneous neutron and plasma irradiation capability

•  Research approach: –  Use of available fission reactor (e.g. High Flux Isotope Reactor/HFIR).

•  Relatively high-flux fast neutron (>0.1 MeV) available to simulate neutron damage •  High-flux thermal neutron accelerates solid transmutation and increases activation

–  Use of thermal neutron-shielding (e.g. cadmium or gadolinium) •  To minimize thermal neutron and simulate fast fusion neutron spectrum.

–  Use of linear plasma device to study sequential neutron-plasma irradiation. •  TPE can handle tritium neutron-irradiated material.

•  Neutron irradiation under US-Japan TITAN (2007-2012) & PHENIX (2013-2018): –  Neutron-irradiation with thermal neutron-shielding at RB* position in HFIR, ORNL

•  500, 800 and 1100 °C irradiation temperature up to 1.0 dpa –  Deuterium plasma exposure in TPE, INL

•  500, 800 and 1100 °C plasma exposure temperature •  > 1022 m-2s-1 D ion flux up to 1027 m-2 D ion fluence

–  Post irradiation examination (PIE) at INL, SNL-NM, ORNL. •  Nuclear reaction analysis (NRA) for D depth profiling •  Thermal desorption spectroscopy for total D retention •  XPS, SAM, PAS, GD-OES for surface/bulk characterization.


Neutron energy spectrum with Gd thermal neutron shielding


1E+09

1E+10

1E+11

1E+12

1E+13

1E+14

1E+15

1E+16

1.E-09 1.E-07 1.E-05 1.E-03 1.E-01 1.E+01

Neu

tron

Flux

, φ

/ cm

-2 s-1

Let

harg

y-1

Neutron Energy, En / eV

JOYO (Mk.III) line6JMTRHFIR (85MW)Fusion FW (RAF Blanket: 1MWn/㎡)

1E+09

1E+10

1E+11

1E+12

1E+13

1E+14

1E+15

1E+16

1.E-09 1.E-07 1.E-05 1.E-03 1.E-01 1.E+01

Neu

tron

Flux

, φ

/ cm

-2 s-1

Let

harg

y-1

Neutron Energy, En / eV

JOYO (Mk.III) line6JMTRHFIR (85MW)Fusion FW (RAF Blanket: 1MWn/㎡)

DEMO (EU) HFIR, Joyo, JMTR

Joyo HP I. Kondo, et al., Nucl. Sci. Technol 18 (1981) 661-472 L.R. Greenwood, J. Nucl. Mater. 216 (1994) 29-44 M.R. Gilbert et al., Nucl. Fus., 52 (2012) 083019.

End of 6 cyclesIrradiation start

SpecimenRegion

RB-19/20J

From 2017 PHENIX SCM

Capsule design, specimen matrix and irradiation schedule

•  RB19J capsule design –  Cost sharing with JAEA-ORNL F82H irradiation program (300 °C zone) –  500, 800, 1100 °C temperature zone for PHENIX tungsten irradiation

•  Specimen matrix for PHENIX Task 3: “Tritium behavior in irradiated tungsten” –  D6TQ : 6 mm OD, 0.25mm thick discs : ~ x10 each temp. zone –  D6TH : 6 mm OD, 0.5 mm thick discs (standard) : ~ x80 each temp. zone –  D6T1 : 6 mm OD, 1.0 mm thick discs : ~ x10 each temp. zone –  D10TQ : 10 mm OD, 0.25 mm thick discs : ~ x15 each temp. zone –  D10TH : 10 mm OD, 0.5 mm thick discs : ~ x10 each temp. zone

•  RB19J irradiation completed –  Started: Cycle 466 (June, 2016) –  Ended: Cycle 469 (December, 2016)


3 RB19J Design and Loading List

Capsule Design Layout

500C 100MM

JAEA

1200C 105MM

800C 110MM

8 RB19J Design and Loading List

D6TQ QTY 22

D6TH QTY 35

D10T1 QTY 10

D10TH QTY 10

D10TQ QTY 15

SS-J QTY 82

D6TH QTY 88

D6T1 QTY 44

TEM QTY 12

6SQ5D QTY 6 NT2

QTY 29

800C Zone Layout

TPE - Tritium Plasma Experiment •  TPE is a unique linear plasma device in four elements:

–  Tritium plasma (< 500 Ci per discharge), –  Divertor-relevant high-flux plasma (>1022 m-2s-1) , –  Moderately radioactive (< 1 mSr/hr @ 30 cm) materials handling –  Beryllium handling

•  Utilizes two containments other than its SS vacuum vessel –  Ventilated enclosure (as a high contamination area/HCA boundary) –  Ventilated PermaCon room (as a contamination area/CA boundary for T)

•  TPE is the only high-flux linear plasma device to investigate tritium retention using tritium and neutron-irradiated material.


Outline of the talk • Motivation and challenges




TPE plasma

Tungsten specimens irradiated under TITAN and PHENIX •  Material and heat treatment:

–  SRW: Stress-relieved (1 h@900 °C in H2 + 0.5 h@900 °C in vacuum) 99.99 wt.% (A.L.M.T., Japan) –  RxW: Recrystallized (1 h@1500 °C + 0.5 h@1300 °C in vacuum) 99.99 wt.% (A.L.M.T., Japan) –  SCW: Single-crystal (100) 99.99 wt.% (Goodfellow, USA)

•  Specimen size: –  Disc: 6 mm dia. 0.2 mm thick (D6TQ) and 6 mm dia. 0.5 mm thick (D6TH) –  Square: 4 x 4 x 0.5 mm3 square (TS) from 16 x 4 x 0.5 mm3 tensile bar specimen


project capsule ID

Damage [dpa] *4

Irradiation temp.[°C ]

sample size

# of sample Irradiated for Task 2-1

shipped to INL?

# of sample available for TPE

PIE completed?

TITAN T9A1 0.025 80 D6TQ 6 (SRW) Yes 4 (SRW) *1 TPE/TDS/NRA

TITAN T9A2 0.3 80 D6TQ 8 (SRW) + 4 (RxW) Yes 2 (SRW) *2 TPE/TDS/NRA

TITAN T9G1 0.3 300 D6TQ 2 (SRW) No 0 *3

TITAN T9G2 2.4 300 D6TQ 4 (SRW) No 0 *3

TITAN TB-300-3 0.1 360 TS Yes 2 (SCW) TPE/TDS



PHENIX PXW-2 0.3 800 D6TH 9 (SRW) Not yet

PHENIX PXW-5 0.3 1100 D6TH 9 (SRW) Not yet

Note: *1) 2 out of 6 were broken during shipment, *2) 10 out of 12 were broken in reactor, *3) all broken or missing *4) 1 dpa ~ 5.0E25 n/m-2 (E>0.1 MeV)

Tungsten specimens exposed to TPE plasma for PMI to date •  Post-Irradiation Examination (PIE) sequence:

–  TPE/TDS: •  TDS (10°C/min to 900°C) performed within 24 hours after the TPE exposure (0.5x1026 m-2)

–  TPE/NRA/TPE/NRA/TDS: •  TDS (10°C/min to 900°C) performed after the 1st TPE exposure (0.5x1026 m-2), 1st NRA, 2nd TPE exposure

(0.5x1026 m-2), and 2nd NRA. The time interval between 1st TPE exposure and TDS were 400-600 days.


capsule ID

Damage [dpa]

HFIR irradiation temp.[°C ]

sample size

sample ID

D/T/He plasma

TPE exposure

temp. [°C ] Ion flux

[1022 m-2s-1] Ion fluence [1026 m-2s-1]

PIE sequence (or status) Ref.

T9A1 0.025 80 D6TQ

Y101 D 200 0.5 0.4 TPE/TDS [1]

Y102 D 100 0.5 0.5+0.5=1.0 TPE/NRA/TPE/NRA/TDS [2] [3] [4]

Y103 D 200 0.6-0.7 0.5+0.5=1.0 TPE/NRA/TPE/NRA/TDS

Y105 D 500 0.7-0.9 0.5+0.7=1.2 TPE/NRA/TPE/NRA/TDS

T9A2 0.3 80 D6TQ Y107 D 200 1.1 0.5+0.5=1.0 TPE/NRA/TPE/NRA/TDS

[3] Y112 D 500 1.1 0.5+0.5=1.0 TPE/NRA/TPE/NRA/TDS

TB-300-3 0.1 760 TS W26A D 700 0.8 0.5 TPE/TDS

W26B D 700 0.8 0.5 TPE/NRA

TB-500-2 0.1 360 TS W53A D 400 0.7 0.5 TPE/TDS

W53B D 400 0.5 0.5 TPE/NRA

TB-650-2 0.1 690 TS W55A D 600 0.8 0.5 TPE/TDS

W55B D 600 0.6 0.5 TPE/NRA Reference: [1] M. Shimada et al., J. Nucl. Mater. 415 (2011) S667, [2] M. Shimada et al., Phys. Scr. T145 (2011) 014051, [3] M. Shimada et al., Nucl. Fusion 55 (2015) 013008, [4] M. Shimada et al., J. Nucl. Mater. 463 (2015) 1005

PIE results from TITAN SRW specimens

M. Shimada (INL) | 38th TFGM | May 9, 2017 | PNNL 13 Reference: [1] M. Shimada et al., Nucl. Fusion 55 (2015) 013008, [2] M. Shimada et al., J. Nucl. Mater. 463 (2015) 1005

Fig.4 in Ref.1

Fig.1 in Ref.2

•  Deep migration of D at elevated temp. [1] –  Specimens:

•  Capsule: T9A1 and T9A2 •  Dose: 0.025 and 0.3 dpa •  HFIR irradiation temp.: 80 °C at HFIR •  TPE exposed temp.: 100, 200, and 500°C

–  NRA results showed that the max. D concentration up to 0.8 at% D/W in 0.3 dpa

–  The large discrepancy between the TDS and NRA, indicating that D was trapped in bulk

•  > 50 µm depth for 0.025 dpa exposed at 500°C •  > 35 µm depth for 0.3 dpa exposed at 500°C

•  Defect annealing of n-irradiated SRW [2] –  Specimens were re-exposed to TPE after TDS

•  TPE condition: 0.5x1026 m-2 D plasma at 200/500 °C •  TDS condition: 10°C/min to 900°C

–  D retention decreases approximately 70% for sample exposed at 500 °C after each annealing,

–  HFIR induced radiation damages were NOT annealed out completely even after the 3rd annealing at 900 °C for 0.5 h (1.5h total)

–  D desorption peak shifts from 800 °C to 600 °C, indicating the change in de-trapping energy and microstructure.

PIE results from PHENIX SCW specimens •  D retention in n-irradiated SCW at DEMO-relevant elevated temp. (1st campaign)

–  Specimens: •  Capsule: TB-300-2, TB-500-2, and TB-650-3 •  Dose: 0.1 dpa •  HFIR irradiation temp.: 360, 690, and 760°C •  TPE exposed temp.: 400, 600, and 700°C

–  TDS (10°C/min to 900°C) performed within 24 hours after the TPE exposure (0.5x1026 m-2) –  1st campaign (TPE/TDS study) showed the competing dynamics of trapping and D diffusion in the bulk

in m-irradiated SCW. •  Diffusion became the dominant transport mechanism, with trapping, or loss of trapping sites from material

annealing, playing a reduced role in the retention process. –  2nd campaign is to measure D depth profile on similar specimens by NRA at SNL-MN (next slide)


Figure: TDS spectra (left) and total retention (right) for neutron-irradiated single crystal W samples (100 eV D with 0.5x1026 D m—2 ) Ref. [1] M. Shimada et al., Nucl. Fusion 55 (2015) 013008,

Depth profile (linear) from HFIR irradiated SCW

•  Nuclear reaction analysis (NRA) by Bill Wampler (SNL-MN)


0.E+00

1.E-03

2.E-03

3.E-03

4.E-03

5.E-03

-0.5 0 0.5 1 1.5 2 2.5 3

D/W

Depth (microns)

unirr W53B W26B W55B

October 2016 W.R.Wampler Sandia National Laboratories Ion Beam Lab

0.E+00

1.E-04

2.E-04

3.E-04

4.E-04

5.E-04

-0.5 0 0.5 1 1.5 2 2.5 3

D/W

Depth (microns)



•  W53B: (THFIR, TTPE)=(360, 400) –  0.3 % D/W near surface < 2 µm and 0.15 % D/W near surface > 2 µm

•  W55B: (THFIR, TTPE)=(690, 600) –  D desorption near surface < 2 µm and 0.015 % D/W near surface > 2µm


1.E-06

1.E-05

1.E-04

1.E-03

1.E-02

-0.5 0 0.5 1 1.5 2 2.5 3

D/W

Depth (microns)



Depth profile (log) from HFIR irradiated SCW

•  Nuclear reaction analysis (NRA) by Bill Wampler (SNL-MN)


•  W53B: (THFIR, TTPE)=(360, 400) –  0.3 % D/W near surface < 2 µm and 0.15 % D/W near surface > 2 µm



Tungsten specimens irradiated under PHENIX •  Material and heat treatment:

–  SRW: Stress-relieved (1 h@900 °C in H2 + 0.5 h@900 °C in vacuum) 99.99 wt.% (A.L.M.T., Japan) –  SCW: Single-crystal [110] (0.5 h@900 °C in vacuum) 99.99 wt.% –  W-3Re: 3%Re/W (prepared by Tohoku U.) and W-5Re: 5%Re/W (prepared by U. of Toyama) –  UFG-W: Ultra fine grained (1.1% TiC)

•  Specimen size: –  6 mm dia. disc: 0.25 mm thick (D6TQ), 0.5 mm thick (D6TH), 1.0 mm thick (D6T1) –  10 mm dia. disc: 0.25 mm thick (D10TQ), 0.5 mm thick (D10TH)


material sample size

# of sample for PHENIX Task 3 PIE plan

500 °C 800 °C 1100 °C

SRW

D6TQ 10 10 10 Plasma-driven D/T/He retention

Gas-driven D retention D6TH 42 42 32 D6T1 10 10 0

D10TQ 15 15 0 Gas-driven T permeation

D10TH 10 10 5 SCW D6TH 10 10 10

Plasma-driven D/T/He retention Gas-driven D retention

W-3Re D6TH 10 10 10 W-5Re D6TH 10 10 10 UFG-W D6TH 5 5 5

Note: All specimens were irradiated at HFIR’s RB* with Gd thermal neutron shielding for 4 cycles.

Neutron-irradiated W specimen exposed to TPE •  Displacement damage vs. irradiation temperature

–  TITAN (’07-’12) focused on low temperature (<500°C) high dose (<< 1 dpa) in pure W –  PHENIX (’13-’18) aims at high temperature (>500°C) high dose (> 1 dpa) in W and W alloy


PHENIX (SRW, SCW, [3-5%]Re/W)

PHENIX (SRW)

TITAN (SRW)

TITAN (SCW)

TITAN (SRW)

PIE plans in PHENIX Task 3 1)  PIE at ORNL (sharing data from Task 2)

–  Positron Annihilation Spectroscopy coincidence lifetime (PAS-CL) for defect characterization

–  Transmission Electron Spectroscopy (TEM) for defect characterization

2)  PIE at STAR, INL –  Plasma-driven D/T/He implantation with TPE for D/T/He implantation –  Gas-driven T permeation with TGAP for T permeation –  Deuterium gas exposure with SGAS for D absorption –  PAS Doppler broadening measurement (PAS-DB) for defect characterization –  Thermal desorption spectroscopy (TDS) for total D/T/He retention measure’t –  X-ray photoelectron spectroscopy (XPS) for surface chemical state –  Scanning Auger Microscopy (SAM) for surface elemental composition –  Glow discharge optical emission spectroscopy (GD-OES) for depth profiling

3)  Nuclear reaction analysis (NRA) at SNL-MN (Bill Wampler)

4)  Supporting PIE at JA Oarai Center –  PAS-DB, PAS-CL, Compact Divertor Plasma Simulator (CDPS), TDS etc.


PIE plans in TPE under PHENIX Task 3 •  TPE experimental condition:

–  ne,max 0.5×1019 m−3 è > 1.0×1019 m−3 (plan *1) –  Γmax 0.5×1023 m−2 s−1 è > 1.0×1023 m−2 s−1 (plan *1) –  qheat < 1.0 MW m−2 è > 1.0 MW m−2 (plan *1) –  Tsample < 700 °C è up to 1200 °C (plan *2) –  Eion up to 600 eV –  Plasma species: He, D, T

•  PIE plans *3 –  D6TH specimens (Task 3 standard specimen)

•  Temperature (500-1100 °C), ion flux (1e22-1e23 m-2s-1), ion fluence (1e26-1e28 m-2) dependence on plasma-driven D/T/He retention

–  D6TQ, D6TH, and D6T1 specimens •  Thickness dependence on plasma-driven D/T/He retention •  Thickness dependence on gas-driven D retention

–  D10TQ and D10TH specimens •  Thickness dependence in gas-driven T permeation

NOTE: *1) New source holder was designed based on UCSD PISCES-A, and is being fabricated at INL. *2) New sample holder is being designed based on UCSD PISCES-A. *3) Actual test matrix will be determined based on the number of intact HFIR irradiated samples after its shipment to INL.


Summary •  Tritium retention in plasma facing components (PFCs) is one of the key

safety issues in magnetic fusion energy.

•  Tritium retention is neutron-irradiated tungsten and tungsten alloys will be predominant factor in DEMO and future fusion reactor.

•  Under US-Japan PHENIX program, the combination of thermal neutron-shielded irradiation at HFIR, ORNL and high-flux plasma exposure at TPE, INL is used to investigate tritium retention in neutron-irradiated tungsten.

•  RB19J irradiation successfully ended in December 2016 with 4 cycles –  Irradiation temperature: 500, 800, 1100 °C –  Radiation damage: up to 1.0 dpa

•  RB19J neutron-irradiated tungsten samples will be shipped to INL in FY18

•  Plan to perform PIE of hundreds of neutron-irradiated samples by March 2019.

•  PIE from RB19J neutron-irradiated tungsten samples will provide valuable database for in-vessel tritium source (i.e. tritium retention) assessment in future DEMO licensing.


Future Directions in FSE Research •  Strategy based on FES guidance and 2013 FES Peer Review Comments

–  Materials Research: Fusion materials, including tungsten irradiated, will be studied at high temperature and heat flux to measure tritium retention and permeation. Dust explosion measurements for fusion materials will continue in support of licensing and computer code development activities. New material diagnostics.

–  Code Development: for the near term, a newer version of MELCOR for ITER will to be completed that includes tritium transport and dust explosion models. Long- term: Multi-dimensional safety code capabilities needs to be developed that take advantage of parallel computing (example RELAP 7)

–  Risk and Licensing: FSP’s evolving failure rate database will be expanded to include maintenance data from existing tokamaks. Risk-informed safety analysis methods (example RISMC Toolkit) will be studied for application to an FNSF. Continue ASME codes and standards and licensing framework development.

–  Collaborations: Participation in existing collaborations to leverage other institution's capabilities and reduce duplication of effort. STAR will move towards being more effective FES User Facility.

22

The National Nuclear Laboratory

M. Shimada (INL) | 38th TFGM | May 9, 2017 | PNNL

Supporting slides


Material challenges in nuclear fusion (2/2)


PMI > 1023 m-2s-1, > 10 MWm-2

•  Erosion & re-deposition •  High heat flux (steady state & transient) •  Surface morphology (tungsten fuzz, 1H/4He bubble)

etc…

neutron >> 1 dpa •  Radiation damage •  Solid transmutation (e.g. Re, Os in W) •  Gas production (1H, 4He) •  Thermal conductivity

degradation •  embrittlement

etc…

tritium D/T

•  Tritium behavior (permeation & retention) •  Tritium decay 3He effect •  Beta-induced effect

etc…

Burning Plasma PMI

samples were irradiated at approximately 500 !C, the precipitatesize in the samples irradiated at approximately 800 !C was largerthan those in the samples irradiated at 500 !C. This differencemight be attributed to the growth of defect clusters at highertemperatures by the higher diffusivity of solute elements.

3.3. Damage structure map of irradiated W

On the basis of our irradiation data and the previously publisheddata obtained by EBR-II irradiation [3e6], damage structure mapsof W and WeRe alloys are suggested, as shown in Fig. 2. Voids arethe major damage structure of pure W for irradiation at tempera-tures greater than 400 !C and at damage levels above 0.17 dpa.Dislocation loops are also observed in lower dpa region, but theirnumber density is less than that of voids. The dashed lines areboundary lines of predicted microstructural regions. For example,the right-hand side of a solid line for a void lattice shows the void-lattice-observed region. The right-hand sides of “5/10Re” dashedline and “26Re” chain line also show the precipitate-observed re-gions of W-5%Re and W-26%Re alloys, respectively. Our datashowed that void lattices were observed in pure W irradiated atlevels as high as 0.96 dpa at 538 !C and at levels as high as 1.54 dpaat 750 !C [13]. The formation mechanism of the void lattice has notyet been clarified [22] but the lattice structure is known to beformed under strict balances of vacancy and interstitial productionand annihilation. In the present work, we considered that as long asthe production and annihilation of point defects in the structure arebalanced and the lattice structure is stable. With increasing irra-diation fluence, transmuted Re increases and it enhances vacancy-

interstitial recombination by Re effects, and then the balance ofproduction and annihilation of defects changes. Finally, the voidlattice structure change. The structure change after the higherirradiation level will be discussed later.

In the irradiated WeRe alloys, acicular precipitates were themajor structure and the precipitation behavior depended on the Recontent and on the damage level. Precipitations in W-26%Re wereobserved at damage levels greater than 0.2 dpa. In W-5/10%Re,precipitates were observed at damage levels greater than 0.5 dpa.In W-3%Re, microstructure data of irradiation sample was notenough to determine the precipitate line.

3.4. Irradiation hardening

The relation between irradiation hardening and the Re contentof starting materials after JMTR, Joyo and HFIR irradiation areshown in Fig. 5 [13]. The irradiation temperature region is400e600 !C for Fig. 5(a) and it is 740e800 !C for Fig. 5(b). Theresults show that at lower dpa levels (lower than 0.5 dpa), theirradiation-induced hardening of Wwas relatively higher than thatof WeRe alloys. The hardening of W was caused by the void anddislocation loop formations, as shown in Fig. 1. The lower irradia-tion hardening of WeRe alloys compared with pure W in this dparange might be attributable to suppression of the formation ofvoids and dislocation loops in WeRe alloys.

The hardening of the WeRe alloys increased after approxi-mately 1 dpa irradiation, as shown in Fig. 5(a) and (b). The results ofmicrostructural observation show that fine and dense precipitatesformed in the hardened specimens and may work as obstacles to

Fig. 4. TEM micrographs of neutron irradiated pure W and W-10%Re alloy at 750 !C in the Joyo reactor ((a) and (b)) and at 800 !C in the HFIR ((c) and (d)).

A. Hasegawa et al. / Journal of Nuclear Materials 471 (2016) 175e183 179

Ref: A. Hasegawa JNM 2016

structure (A) and ring-like structure (B), see Fig. 1c). In these pro-cesses, growth of the highly pressurized He bubbles in the surfaceregion could lead to formation of protrusions, the size of which de-pends on the size of bubbles. Moreover, blister-like structures canform in some parts by the development of the He bubble growth.It is thought that active surface diffusion of tungsten atoms playsa key role in forming protrusions and fine structures, becausesurface diffusion is, in general, much faster than lattice diffusion.It is likely that the surface diffusion allows W to combine with Hebubbles to change the surface morphology.

The nature of the growth of nano-structured layers with the Heflux of (4–6) ! 1022 m"2 s"1 at 1120 K in the PISCES linear plasmadevice is revealed by the scanning electron microscope (SEM)images in Fig. 2 [5]. As the exposure time is increased, the nano-structured layer grows in thickness to more than 5 lm for the timeof 2.2 ! 104 s. Layer thickness data are shown for exposure tem-perature regimes of 1120 and 1320 K. For both exposure tempera-tures, the data are well characterized by a square root dependenceon time, suggesting the diffusion-like process dominates nano-structure growth. Similar dependence was also seen in the NAGDISlinear plasma device with an incubation He fluence of about4 ! 1024 m"2 [14]. In addition, formation of the nano-structure ishardly affected by simultaneous D implantation [15]. Althoughthe growth mechanism of the nano-structure is not well under-stood, a model called ‘‘Visco-elastic model’’, which predicts thet1/2 growth of the nano-structure, was recently proposed by Kra-sheninnikov [16].

Formation conditions for the nano-structure were reported byKajita et al. [13]. The incident ion energy and surface temperatureare key parameters for the formation. Fig. 3 shows the formationwindow of the nano-structure. The gray colored region and thehatched region correspond to the conditions in which nano-struc-ture and large helium bubbles, respectively, are observed after he-lium irradiation. The formation window ranges between 1000 Kand 2000 K, and over the ion energy of about 20 eV. The thicknessof the nano-structure tends to increase with temperature. The en-ergy threshold is #12 eV for the nano-structure formation on VPS(vacuum vapor deposition) W on graphite [4]. This threshold en-ergy is somewhat higher than the surface barrier potential of Wfor He (#6 eV) [3]. The reason why the energy threshold for thenano-structure formation (12–20 eV) is higher than the barrier po-tential is not known. The formation rate of the nano-structure maybe too low for detection if the energy lies below the thresholdenergies.

In ITER, heat and particle flux at the outer strike point would beabout 10 MW m"2 and 1024 m"2 s"1 [17]. In NAGDIS and PISCES,the highest He flux is about 1023 m"2 s"1, which would be aboutthe same as the highest He flux to divertor plates in ITER, becauseHe concentration in burning plasmas would be 5–10%. The heatflux to W specimens in these linear plasma devices, however, isabout an order of magnitude lower than to ITER. He irradiationwith comparable heat and particle flux to ITER divertor has been

Fig. 1. He irradiated W in NAGDIS-II at 1400 K and 50 eV. He fluences are (a), (b)1.1 ! 1025 m"2 and (c) 2.4 ! 1025 m"2 [14].

Fig. 2. Cross-sectional SEM images of W targets exposed to pure He plasma at 1120 K for exposure times from (a) 300 s to (e) 22,000 s with the He flux of (4–6) ! 1022

m"2 s"1 [5].

S268 Y. Ueda et al. / Journal of Nuclear Materials 442 (2013) S267–S272

Ref: M.J. Baldwin NF 2009

5

Tritium Embrittlement Process

GRAPHIC

Decay helium precipitate as nanometer-sized bubbles that harden the microstructure

Embrittlement and Cracking of Grain

Boundaries

Ref: M.Tosten & M. Morgan WSRC-RP-92-551 (1998)

sample #: W12e shot #: 252 ion energy: 100 eV sample current: 2 A temperature: 230 °C duration: 1 hr

The large features below are due to grain ejection from the surface.

An intermediate-range image shows bubbles only within select grains. The grain size in the ITER-grade W appears to be on the order of 50 microns.

The bubble size is much smaller than observed on PLANSEE W (~ 5 microns)

ITER%grade%W%(A.L.M.T.)%

Hot5rolled%W%(Plansee)% W%mel;ng%(by%arc)%

Ref: M. Shimada TPE

Irradiation temperature analysis


Sub-capsule

Rholder (mm)

Holder Material

Spacer Material

T_avg (°C)

250 Top 16.17 Aluminum Aluminum 290

250 Mid 16.17 Aluminum Aluminum 238

250 Btm 16.17 Aluminum Aluminum 220

500 16.15 Graphite Graphite 482



Upgrade in plasma diagnostics •  Optical spectrometer was installed in FY16

–  Specification: •  0.75m focus length Czerny-Turner spectrograph (Andor Shamrock 750) •  1024x255 CCD camera (Andor iDus420BU)

–  Capability: •  Provide plasma impurity measurement •  He ion concentration measurement •  Neutral gas temperature measurement


•  First mixed-plasma (He + D) was achieved –  In 30 year TPE history –  He I emission lines was observed in newly

setup optical spectrometer system –  Subsequently desorbing helium atom was

observed at elevated temperature during thermal desorption

–  Calibration of CCD camera is required for quantitative measurement of He ion conc.

•  Next step is to produce tungsten fuzz.

New surface diagnostics at STAR •  Glow discharge optical emission spectroscopy

–  Quantitative elemental depth profile analysis –  Nanometer resolution –  Can analyze 10s or even 100s µm sample depth –  Installation completed in FY16

•  X-ray photoelectron spectroscopy –  Surface sensitive chemical analysis –  Excellent for quantification –  Particularly useful for surface effects related to permeation –  Arriving FY17

•  Scanning Auger electron spectroscopy –  AES provides elemental characterization –  Scanning mode allows for microscopy –  Arriving FY17

NOTE: all three diagnostics will be capable of handling low activation and tritium-exposed materials.


New additions to STAR

10

NIMIIX(footprint)

To103

To101

Workbench

LN2

Proposedconfigura-on

Minfloorspacerequired(7’x11’)

XPS

ToolboxGD-O

ESDesk

chiller

I&CRack Desk

Vacuumconsole

ElectronicsRackRoughpump

Minfloorspacerequired(9’x13’)

AES

Workben

ch

I&C

Rack

PAS

20

Glow discharge optical emission spectrometry (GD-

OES)

Coincidence positron annihilation spectroscopy

(CDB-PAS)

X-ray photoelectron spectroscopy (XPS)

Scanning auger microprobe (SAM/SEM) M. Shimada (INL) | 38th TFGM | May 9, 2017 | PNNL 28

Glow discharge optical emission spectroscopy (installation completed in FY16)

29

•  Technique –  Elemental characterization with depth

profiling. –  Analyzed 10s of elements

simultaneously. –  Nano-meter resolution. –  Sputtering rates of nm/sec.

•  Specifications –  Capable of differentiating H and D. –  Realtime depth measurements. –  Monochromator can be used to look at

unknown element. •  GD-OES will measured the D depth

profiling from bulk (>> 10µm) neutron-irradiated tungsten exposed to plasma

One of very few surface analysis techniques capable of measuring hydrogen. 4 mm dia “burn”

M. Shimada (INL) | 38th TFGM | May 9, 2017 | PNNL

Glow discharge optical emission spectroscopy

•  The glow discharge burns a crater into a sample. – Crater shown is 4 mm diameter. – Crater size can vary from 1-10 mm with interchangeable

anodes. •  Fast sputtering rate (over 2 µm/min).

– Material dependent. – Analyze 100s µm in minutes.

Tungsten disk (1 in dia.)


X-ray photoelectron microscopy (XPS) and Scanning Auger electron spectro/microscopy (SAM)

31

•  XPS –  Technique

•  Excellent chemical sensitivity. •  Expansive libraries. •  Capable of detecting elements, Li and

larger. –  Specifications

•  X-ray monochromator for high resolution XPS scans.

•  Multiple x-ray sources. •  Sputtering ion gun for depth profiling.

•  SAM –  Technique

•  Excellent elemental sensitivity. •  Limited quantification. •  Rastering electron beam + Secondary

electron detector = SEM –  Specifications

•  LaB6 filament. •  Sputtering ion gun for depth profiling.

•  XPS and SAM will be installed in FY17 M. Shimada (INL) | 38th TFGM | May 9, 2017 | PNNL

Tritium retention in plasma-facing components 9 - Shimada - Tritium...Masashi Shimada Fusion Safety...

Documents

Transcript of Tritium retention in plasma-facing components 9 - Shimada - Tritium...Masashi Shimada Fusion Safety...