Ion-Driven Permeation of Deuterium through Tungsten Motivation Permeation experiment Results Next steps A. V. Golubeva, M. Mayer, J. Roth.

Ion-Driven Permeation of Deuteriumthrough Tungsten

• Motivation

• Permeation experiment

• Results

• Next steps

A. V. Golubeva, M. Mayer, J. Roth

Ion-driven permeation (IDP)• Hydrogen recycling

• Deep diffusion Hydrogen inventory in the bulk,

cooling water channels

Motivation

Motivation (2)

In this work:

• A new PERMEX set-up for investigation of IDP through metal (W) membranes

• Permeation experiments with W foils

• Influence of surface impurities on IDP through W foils

• Lack of data on ion-driven permeation through W

• No data on ion-driven permeation through coated W

Experimental Temperatures 22-750 0CParticles D2

+ or D3+

Energies 200 eV – 3 keV/DFlux: 1017 – 1018 D/m2sNormal incidenceBackground <510-9 mbar

Ion gun

PERMEX set-up

Registration of HD (main component for W)D2 and other masses - by QMS

QMS calibration by set of calibrated D2 leaks

Membrane backside cleaning– by 1.5 keV Ar+, 51017 Ar/m2s

Calibrated D2

leaks

Author Anderl Nakamura Present work

Т, К 610-823 550-770 < 813-1023

Incident particles H, D D, T D

E, keV/particle 1 1 0.2 – 3

Flux, part./m2s 6.51019 51018 - 1019 1017 - 1018

Backside cleaning No No Yes

Membr. thickness 25 µm, 50 µm0.1 mm/SS

25 µm, 50 µm 50 µm – 0.3 mm

Materials PCW, W coatings on SS

PCW PCW

Measured experimental data

Shapes of permeation curvesPerm. flux (Fperm ) via material of membrane

Lag time (Т)Fperm (Т)Fperm(F0)Isotope effect

Fperm (Т)Lag time (Т)Fperm (surface conditions)

Modelling Energy & Concentr. of traps

Trapping energy Recombination coefficients

Comparison with other IDP W experiments

Materials investigated - 99.98% W foils

NRA on top of both as received W foils are presentO 2.5∙1016 O/cm2 3.6 nm WO2 (and comparableC 3∙1016 C/cm2 thickness of C)

50 µm (Unknown),Not pre-annealed

SEM: grains up to 40 µm SEM: grains 1 – 5 µm SEM: grains 1 – 5 µm

50 µm (Plansee),Not pre-annealed

0.3 mm (Goodfellow),pre-annealed (1500 K, 3 hours)

Permeation curve without front side cleaning

1000 2000 3000 4000

0,0

2,0x1010

4,0x1010

6,0x1010

8,0x1010

1,0x1011

1,2x1011

1,4x1011

1,6x1011

dN

/dt,

D/s

t, s

2,8E13 D/s

Typical permeation curve at first irradiation, T=700 C, material without specification

Spike (due to oxide layer on front side) presents at T>7000Conly at first implantation

D (200 eV) => W 50 µm

T=7000C

Permeation curves &Typical times

0 1 2 3 4 5 60,000

0,001

0,002

0,003

0,004

0,005

0,006

0,007

0,008

0,009

Fpe

rm/F

impl

ante

d

Time, hours

Temperature time to reachsteady-state

700 K 1 h 550 K >10 h

20 min – time to remove impurities from front side by sputtering(Note: D (200 eV) does not create displacement defects in W)

Fimplanted =F0*(1-Rn)

Reflection coefficient Rn – from modeling

Permeating flux / implanted flux 50 µm not annealed foil

820 840 860 880 900 920 940 960 9800.000

0.001

0.002

0.003

0.004

0.005

0.006

Dirty 1 hour cleaning of backside

I perm

/I 0

T, K

Influence of backside cleaning

Backside was notcleaned

Backside was cleaned

An order of magnitude increase after removal of oxide layer at outlet side

D on 50 µm W

Reproducibility of results

Backside was not cleanedbetween implantations

Backside was cleanedbetween implantations

0 2 4 6 8 10 120,000

0,001

0,002

0,003

0,004

0,005

Pe

rme

atio

n r

ate

t, hours

1st implantation 2nd implantation

D(200 eV) =>W, 50µm T=6000C

0,000

0,001

0,002

0,003

0,004

0,005D=>W, 50µm T=6000C

54320 1

Pe

rme

atio

n r

ate

t, hour

1st impl. 200 eV/D 2nd impl 200 eV/D 3rd impl. 200 eV/D impl.1200 eV/D

•Shape (e.g. „Lag time“) – repeatable•Maximum decreases from implantation to implantation

Reproducible (both “lag time” and amplitude of curve)

D on W, 50 µm, 600 0C, 1018 D/m2

Backside was cleaned after sample installation

From sample to sample – 20 % difference in permeation rate

750 800 850 900 950 10000.0000

0.0005

0.0010

0.0015

0.0020

0.0025

as received backside

cleaned

Per

mea

ting

flux

/ In

cide

nt f

lux

Temperature, K

a

b1

b2

b3

Oxide influence: Backside cleaning

Plansee 50 m W, 200 eV/D, 21017 D/m2s Backside cleaning 5 times increase of permeation flux

b3 Surface conditions do not change during at least 2 days

Permeation dependsstrongly on surfaceconditions

2E-3

4E-4

Repeatability from sample to sample – 30 %

Nakamura: permeating flux is proportional to incident flux, F=kF0

in the range 2.5∙1018 – 1019 D/m2s

PERMEX: F=kF0 in the range 5∙1016-2∙1017 D/m2s

To compare results, we consider permeation rate F/F0

500 550 600 650 850 900 9500.0000

0.0005

0.0010

0.0015

0.0020

0.0025

Nakamura for 25 µm W, 1 keV/DPERMEX 50 µm W, 200 eV/D

as received backside cleaned

Pe

rme

atin

g f

lux/

Inci

de

nt

flux

Temperature, K

a

b1

b2

b3

Previously defined Permeation rates through Wcan be underestimated

Oxide influence: Backside cleaning

2003

Database

500 600 700 800 900 10000.0000

0.0005

0.0010

0.0015

0.0020

0.0025

0.0030

0.0035

0.0040

0.0045 Nakamura, 25 µm W

? W, 50 µm

3000 eV, NOT cleaned 200 eV, NOT cleaned 200 eV, CLEANED

Plansee 50 µm W 200 eV, one sample, NOT cleaned 200 eV, one sample, backside cleaned different samples, backside cleaned

Plansee 0.3 mm W 0.3 mm Plansee W, 200 eV

Fpe

rm/F

0

Temperature, K

Permeation strongly depends on material structure (2 times difference for 99.98% W foils of the same thickness but different grain sizes)

Possibility of experiments with thick (0.3 mm) membranes is demonstrated

Grains ~ membrane thickness

Effective diffusion coefficients

• Large influence of traps

0,4 0,6 0,8 1,0 1,2 1,4 1,6 1,8 2,0 2,2

1E-14

1E-13

1E-12

1E-11

1E-10

1E-9

1E-8

1E-7

Effe

ctiv

e di

fusi

on c

oeffi

cien

t D, m

2 /с

1000/T, K-1

Gas interaction with W Frauenfelder, experimental data Frauenfelder, approximation Zaharov, experimental data Zaharov, aproximation Esteban Franzen

Ion Driven Permeation Nakamura, experimental data

present work 0.3 mm W-Gf, pre-annealed (3 h 1500 К) 50 µm W-un, not annealed 50 µm W-Pl, not annealed

2

6

1 LDeff L – foil thickness, - lag time

Conclusions

• PERMEX set-up developed and build allows IDP experiments with control of both membrane sides

• First D on W permeation data obtained for 550-7500C

• 200 eV/D permeation ratio (Fperm/F0) = 5×10-4 (Plansee 50 µm, as received)

• Permeation spike observed at 7000C sputtering of oxide layer on the inlet surface

• Surface layer conditions strongly influence permeation rate: backside cleaning increases permeation by factor 5

• Material structure influences significantly IDP

Future plans

• Permeation through W with different coatings on front side (oxide, carbide)

PhD thesis in collaboration with MEPhI (A. Pisarev)

• Modeling of IDP to obtain recombination coefficients