Recrystallization behaviour of high-flux hydrogen plasmaexposed tungstenCitation for published version (APA):Shah, V., Beune, J. T. S., Li, Y., Loewenhoff, T., Wirtz, M., Morgan, T. W., & van Dommelen, J. A. W. H. (2021).Recrystallization behaviour of high-flux hydrogen plasma exposed tungsten. Journal of Nuclear Materials, 545,[152748]. https://doi.org/10.1016/j.jnucmat.2020.152748

Document license:CC BY

DOI:10.1016/j.jnucmat.2020.152748

Document status and date:Published: 01/03/2021

Document Version:Publisher’s PDF, also known as Version of Record (includes final page, issue and volume numbers)

Please check the document version of this publication:

• A submitted manuscript is the version of the article upon submission and before peer-review. There can beimportant differences between the submitted version and the official published version of record. Peopleinterested in the research are advised to contact the author for the final version of the publication, or visit theDOI to the publisher's website.• The final author version and the galley proof are versions of the publication after peer review.• The final published version features the final layout of the paper including the volume, issue and pagenumbers.Link to publication

General rightsCopyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright ownersand it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.

• Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portal.

If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above, pleasefollow below link for the End User Agreement:www.tue.nl/taverne

Take down policyIf you believe that this document breaches copyright please contact us at:openaccess@tue.nlproviding details and we will investigate your claim.

Download date: 22. Dec. 2021

Journal of Nuclear Materials 545 (2021) 152748

Contents lists available at ScienceDirect

Journal of Nuclear Materials

journal homepage: www.elsevier.com/locate/jnucmat

Recrystallization behaviour of high-flux hydrogen plasma exposed

tungsten

V. Shah

a , J.T.S. Beune

a , Y. Li a , b , Th. Loewenhoff c , M. Wirtz

c , T.W. Morgan

b , J.A.W. van

Dommelen

a , ∗

a Department of Mechanical Engineering, Eindhoven University of Technology, PO Box 513, Eindhoven, 5600 MB, the Netherlands b DIFFER - Dutch Institute for Fundamental Energy Research, De Zaale 20, Eindhoven, 5612 AJ, the Netherlands c Forschungszentrum Jülich, Institut für Energie- und Klimaforschung, Jülich 52425, Germany

a r t i c l e i n f o

Article history:

Received 22 October 2020

Revised 26 November 2020

Accepted 12 December 2020

Available online 17 December 2020

Keywords:

Tungsten

Hydrogen

Plasma exposure

Hardness

Recrystallization kinetics

a b s t r a c t

Knowledge of a material’s thermal stability under extreme synergistic particle and heat loads is crucial for

developing high performance reactor materials. In this work, the recrystallization behaviour of tungsten

under the influence of hydrogen is investigated by low energy high flux hydrogen plasma exposure for

various lengths of time. The microstructural changes following exposure are probed by micro-indentation,

electron back-scatter diffraction measurements and the characteristic time for recrystallization is assessed

using the Johnson–Mehl–Avrami–Kolmogorov (JMAK) model. A recrystallization activation energy in the

range of 425 to 440 kJ.mol -1 is determined, identical to that of oven annealed samples, thereby indicating

an insignificant influence of hydrogen plasma on the recrystallization kinetics of tungsten.

© 2020 The Author(s). Published by Elsevier B.V.

This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ )

1

b

f

d

t

f

o

[

t

l

e

p

r

o

c

t

l

p

b

e

o

P

b

m

r

fi

f

o

t

f

i

d

i

t

c

t

o

d

l

h

0

. Introduction

Over the past decade, the refractory metal Tungsten (W) has

een identified as the leading candidate material for the plasma

acing components (PFCs) for the divertor of future fusion reactors

ue to its favourable high temperature properties. The lifetime of

hese W PFCs is crucial for the efficiency, economics and good per-

ormance of the reactor as the divertor performs the vital function

f extracting heat, helium and other impurities from the plasma

1,2] . The PFCs are subjected to intense particle fluxes of fast neu-

rons, low energy hydrogen (H) and helium (He) ions and radiation,

eading to high heat loads ( ∼10 to 20 MWm

-2 along with high en-

rgy transients) [3,4] . Fundamentally speaking, the prolonged ex-

osure of metals at high temperatures assists in microstructural

estoration by annihilation of defects, nucleation and the growth

f new defect free grains to restore the equilibrium state; a process

ommonly termed as recrystallization and grain growth [5] . Under

he expected loading conditions in the ITER tokamak, the recrystal-

ization of at least top few mm of the ITER bulk-W PFCs is antici-

ated [6] . Recrystallization of W during high heat flux loading has

een observed experimentally by several studies following laser,

∗ Corresponding author.

E-mail address: j.a.w.v.dommelen@tue.nl (J.A.W. van Dommelen).

a

o

i

w

ttps://doi.org/10.1016/j.jnucmat.2020.152748

022-3115/© 2020 The Author(s). Published by Elsevier B.V. This is an open access article

lectron, as well as plasma exposure [7–10] . The recrystallization

f bulk W has been surmised as detrimental for the lifetime of the

FCs due to a loss of strength and hardness, and an increase in the

rittle-to-ductile transition temperature (BDTT), consequently pro-

oting crack formation and adversely affecting the thermal fatigue

esistance of the PFCs [11–13] . Since, the “ITER grade” W is de-

ned within a narrow range of parameters produced by heavy de-

ormation, the occurrence of recrystallization also implies the loss

f the designed benefits of the purposely engineered microstruc-

ure of W. However, it should be noted that the picture is not

ully complete, as there have been studies depicting an increase

n the total elongation following recrystallization as compared to

eformed state (similar testing temperature), implying a decrease

n BDTT [11,14,15] .

Besides the heat loads, the influence of particles loads, i.e. neu-

rons and plasma ions, on the kinetics of the recrystallization pro-

ess at the microstructural scale are not well defined. Primarily,

he expected consequence of neutron irradiation in W PFCs is an

verall increase in the internal stored energy because of the high

ensity of defects created (vacancies/ self interstitial clusters, dis-

ocation loops as well as network dislocations), thereby leading to

n increase in the recrystallization rate as compared to the thermal

nly case [16–18] . Also, the synergistic interaction of the neutron

nduced defects with plasma species, specifically the He atoms,

ill assist in the formation of gas filled voids, frequently termed

under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ )

V. Shah, J.T.S. Beune, Y. Li et al. Journal of Nuclear Materials 545 (2021) 152748

H

i

e

d

t

l

p

a

b

c

d

T

e

t

t

b

p

a

(

i

t

o

t

t

t

d

e

b

t

v

i

r

2

2

d

t

t

t

p

e

e

2

t

fi

2

u

w

u

i

s

t

s

e

m

s

i

m

p

w

t

a

h

a

s

s

p

d

i

l

p

v

b

a

c

2

d

i

d

p

o

o

t

f

F

p

t

t

s

F

m

a

e bubbles [19] . In general, the effect of bubbles on the kinet-

cs of the recrystallization process have been well known [20] . For

xample, doping of W with potassium (K) has been practised for

ecades to retard recrystallization by producing K bubbles, even-

ually resulting in microstructural strengthening [21,22] . A simi-

ar effect of retarding recrystallization by He bubbles has been re-

orted under high energy He ion irradiation of W [23–25] . Mech-

nistically, the impediment of recrystallization in the presence of

ubbles is similar to the effect of second phase particles (pre-

ipitates), with the bubbles/particles exerting a drag force (Zener

rag) on the moving grain boundaries during recrystallization [5] .

he magnitude of the drag force, in the case of bubbles, is gov-

rned by the size of the bubble, showing a further dependency on

he irradiation/annealing temperature [25] . In the case of H atoms,

he identical conclusion on the role of gas atoms has been drawn

y Loewenhoff et al. [26] , where recrystallization of W was sup-

ressed during H plasma exposure regardless of the high temper-

tures. Contrarily, the presence of dissolved H atoms in Palladium

Pd) and Vanadium (V) has been reported to accelerate the kinet-

cs of the recrystallization and grain growth process by lowering

he formation energy of vacancies along with a change in mobility

f dislocations and other defects [27,28] . Thus, given the inconsis-

ent conclusions from these studies, the fundamental question on

he role of H atoms as expediter or inhibitor of the recrystalliza-

ion process ensues and specifically for the fusion reactor case, un-

erstanding its influence is critical for obtaining accurate lifetime

stimates of the W PFCs. The present work addresses this question

y using the unique high-flux linear plasma facility Magnum-PSI

o perform synergistic high heat and H particle exposure of W for

arying time intervals at temperature ranges foreseen in PFCs. This

s followed by a systematic post-mortem analysis to determine the

ecrystallization kinetics and compared with literature studies.

. Experimental

.1. Test samples

A tungsten bar, manufactured by Plansee using sintering and bi-

irectional forging was used as the source for the samples. These

op hat shaped samples had an upper surface of 30 × 30 mm and a

hickness of 10 mm ( Fig. 1 ), and were produced at Forschungszen-

rum Jülich using electro-discharge machining. The machined sam-

les had elongated grains with an average aspect ratio of 4.5, ori-

nted perpendicular to the plasma exposure plane, while the av-

rage grain size (equivalent circular diameter) was approximately

1 μm. The bi-directional forging based manufacturing of the ma-

erial induced a strong 〈 110 〉 type fibre texture (pre-dominantly αbre, see Appendix A ), alike to a wire drawing process [29] .

ig. 1. Schematic showing the geometry of the samples employed for the H plasma expos

illimetres. The particle flux in the beam has a Gaussian like distribution (indicated in (a

long the exposed surface as shown in (b) .

2

.2. Magnum PSI: hydrogen plasma exposure

The H plasma exposure was carried out in a continuous mode

sing the linear plasma device Magnum-PSI [30] . This is equipped

ith superconducting magnets and capable of generating contin-

ous ion fluxes of the order 10 24 to 10 25 m

-2 s -1 as anticipated

n future fusion reactors [31,32] . For the investigation, a total of 5

amples were employed, with exposure times varying from 10 min

o 6 hr. The plasma conditions were monitored using Thomson

cattering (TS) and gave electron temperatures (T e ) 2 to 3 eV and

lectron densities (n e ) of 10 19 to 10 20 m

-3 . During exposure, the

achine settings were constantly monitored to maintain a peak

urface temperature of 1500 °C. The targets were exposed at float-

ng potential and following sheath acceleration, this led to a maxi-

um ion energy of 14 eV at the target surface, corresponding to a

rojected implantation range of approximately 1 nm. This energy is

ell below the displacement threshold energy (85 to 90 eV [33] ),

hereby assuring no generation of vacancies. The plasma beam had

Gaussian particle and heat flux distribution with a full width

alf maximum (FWHM) of approximately 12 ± 1 mm, resulting in

non-uniform temperature distribution radially over the exposed

urface as shown in Fig. 1 b. The sample surface temperature at the

ample centre was continuously tracked using a multi-wavelength

yrometer (FAR Associates® FMPI) with a spot size of ≈ 3 mm. Ad-

itionally, the entire surface was monitored every 8 min via a fast

nfrared (IR) camera (FLIR SC7500MB). Here, due to the extreme

ength of some of the exposures, continuous recording was not

ossible. The sample temperature was also continuously tracked

ia an N-type thermocouple mounted in a hole machined at the

ottom of the sample ( Fig. 1 ). This gave values between 600 °C

nd 800 °C for all samples, with variations due to differences in

lamping of the target samples to the water cooling system.

.3. Post exposure characterization

Post plasma exposure, insights into the microstructural changes

ue to recrystallization were probed by performing micro-

ndentations, as recrystallization is characterized by a distinct re-

uction in hardness. The micro-indentation measurements were

erformed using a CSM micro-indenter with an indentation load

f 4.9 N (HV0.5). For better statistics and to scrutinize the effect

f the non-uniform temperature distribution on the microstruc-

ure evolution of the exposed surface, the indentations were per-

ormed in a mapping mode with a minimum step size of 200 μm.

or selected samples, the indentation measurements were com-

lemented with electron back-scatter diffraction (EBSD) mapping,

hereby allowing to correlate the indentations with the microstruc-

ural state. The EBSD based mapping along the plasma exposed

urface was performed using a TESCAN-MIRA SEM with an acceler-

ure (a) front view (b) isometric view (not to scale). All the given dimensions are in

) ) with a FWHM of 12 ± 1 mm, leading to a non-uniform temperature distribution

V. Shah, J.T.S. Beune, Y. Li et al. Journal of Nuclear Materials 545 (2021) 152748

a

o

3

f

m

p

t

c

a

t

s

a

n

a

s

t

g

a

e

v

e

m

fi

H

s

i

t

s

t

t

v

i

c

d

v

l

a

b

i

m

c

r

t

r

F

r

w

f

tion voltage of 20 kV and a step size of 0.25 to 3.5 μm, depending

n the microstructural state.

. Results and discussion

The temperature distribution along the exposure plane obtained

rom the IR recordings for varying exposure times are shown as 2D

aps in Fig. 2 a. An exception is the 6 hr exposure, where the tem-

erature map was estimated based on the plasma beam parame-

ers (FWHM = 13 mm determined using Thompson scattering) and

onstrained by the pyrometer data due to erroneous IR recording

rising from copper deposition from the source and related con-

amination of the IR measurement. The influence of copper depo-

ition on the hardness measurements is expected to be negligible,

s the deposited copper layer has a thickness of not more than few

anometres. As seen in the temperature maps ( Fig. 2 a), small vari-

tions in the alignment of the plasma during exposure between

amples manifests as a shift in the position of the hot zone from

he centre of the sample. Additionally, variations in temperature

radient along the exposure plane, i.e. away from the beam spot

nd towards the sample edge can be observed between different

xposures (for example: 10 min vs. 3 hr). This can be attributed to

ariation in the heat transfer capacity of the clamping between the

xposures.

ig. 2. Exposure time dependent temperature and hardness maps (a,b) , along with EBSD

egion near the centre of the sample following 1 hr H plasma exposure. The black boxes

ere performed as shown in (c) and in Appendix B . A radial recrystallization extent of

ollowing 1 hr exposure (in (c) and (d) ). (For interpretation of the references to colour in

3

Fig. 2 b shows the exposure time dependent hardness (HV)

aps acquired from the micro-indentation measurements. At a

rst glance, independent of the exposure time, significantly lower

V values can be observed in the areas close to the centre of the

ample, while the area with low HV values increases with increas-

ng time (for example: between 1 hr and 3 hr). It is worthwhile

o note that for lower exposure time (10 min to 1 hr), the above

tatement does not hold entirely. On further close comparison of

he temperature and the HV maps ( Fig. 2 a and Fig. 2 b), a dis-

inctive relationship between the high temperature and low HV

alue (vice versa) regions can be realized. Thereby, clearly depict-

ng the recrystallization-induced softening due to microstructural

hanges during exposure. Characteristically, all the hardness maps

ominantly display sharp transitions between the low and high HV

alue zones. These transition regions correspond to the relatively

ower recovered/partially recrystallized fractions as higher temper-

tures favour recrystallization. Fig. 2 c shows a large scale EBSD

ased mapping of the recrystallized area following 1 hr exposure

n terms of a inverse pole figure (Z-IPF) map. The recrystallized

icrostructure (dominantly with preferred orientation [101], green

oloured) clearly follows the radial temperature profile, with the

ecrystallization extent radially extending up to ≈ 4.4 mm along

he exposure plane from the hot-spot centre, conforming to the

adii of low HV zone in Fig. 2 b (black box in 1 hr HV map). To fur-

based IPF map of the Z-direction (depth) (c) and GOS map (d) of the recrystallized

in 1 hr and 3 hr HV maps (b) correspond to the areas where EBSD measurements

approximately 4.4 mm from the hot-spot centre (orange colour circle) is observed

this figure legend, the reader is referred to the web version of this article.)

V. Shah, J.T.S. Beune, Y. Li et al. Journal of Nuclear Materials 545 (2021) 152748

Fig. 3. (a) The evolution of hardness with respect to temperature for different exposure time and their corresponding parametric functions (black line) used for determining

the time dependent temperature at which half-hardness loss occurs ( t hal f ). (b) The temperature dependent temporal evolution of the recrystallized fraction. The solid lines

represent the evolution at different temperatures based on the JMAK model, while the circular markers illustrate the recrystallized fraction data determined from hardness

following different exposure times at 1350 °C.(For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

t

g

s

t

t

i

m

d

A

fi

a

l

t

n

h

d

a

f

(

m

H

a

s

T

d

a

t

p

(

t

f

t

g

o

1

t

t

f

u

(

t

t

a

s

a

t

t

(

f

Table 1

Exposure time dependent temperature for half-hardness

loss obtained from Fig. 3 a.

T [ °C] t hal f

1399 ± 9.4 10 min

1371 ± 3.8 30 min

1366 ± 4.8 1 hr

1255 ± 3.3 3 hr

1247 ± 8.6 6 hr

her ascertain the recrystallized state in Fig. 2 c, Fig. 2 d presents the

rain scale misorientation analysis in terms of a grain orientation

pread (GOS) map. The majority of grains (99.93% area fraction)

end to have an orientation spread below 3 °, which is typically the

hreshold value for discriminating recrystallized grains especially

n W [9,15] .

The HV values extracted from the varying time dependent HV

aps ( Fig. 2 b) and correlated with the corresponding temperature

istribution ( Fig. 2 a) along the exposure plane are shown in Fig. 3 a.

s highlighted before, the temperature dependent hardness pro-

le depicts a two stage behaviour, where small drops in HV values

t low temperatures occur pre-dominantly due to recovery, while

arge drops in HV occur at high temperatures pre-dominantly due

o recrystallization. It is important to point out that for some hard-

ess measurements, the deviations tend to be large (for example,

igh HV values for 1 hr exposure). Nonetheless, the temperature

ependent HV trend for all exposure times displays an identical

nd consistent behaviour. The temperature dependent parametric

unctions determined from the varying exposure time data sets

Fig. 3 a) are further shown as black lines in Fig. 3 a. Here, the para-

etric function represents an error type function having the form

V = 0.5 (1 − erf (T − a )/ b ) c + d , with HV denoting the hardness

nd T denoting the temperature; while a , b , c and d are con-

tants determined by non-linear least square regression procedure.

he scattering of hardness data at lower temperatures prevents a

efinite analysis of the kinetics in terms of the combined recovery

nd recrystallization processes. Thus, the recovery kinetics are not

reated explicitly in the following analysis.

For assessing the recrystallization kinetics during exposure, the

arameter t half , representing the time scale for half hardness-loss

or half recrystallization), is introduced. This parameter defines the

4

hermal stability of the deformed microstructure, and is obtained

rom the parametric functions (black line) shown in Fig. 3 a by de-

ermining the temperature at which half hardness-loss occurs for a

iven exposure time. Table 1 underlines the temperature evolution

f t hal f , and an apparent drop of nearly 150°C can be seen between

0 min (1400 °C) and 6 hr (1250 °C) exposure, implying a reduced

hermal stability at longer exposure times.

To assess the recrystallization kinetics over a wide tempera-

ure range, the temperature dependent recrystallized fraction ( X )

or a given exposure time are determined from the mean HV val-

es ( Fig. 3 a). Next, the modified Johnson-Mehl-Avrami-Kolmogorov

JMAK) model having the form X = 1 − exp ( −b n (t − t inc ) n ) is used

o determine the kinetics [5] . Here, the coefficient b represents

he thermal activation, the Avrami exponent n describes the nucle-

tion mechanism including the growth dimensionality, t inc repre-

ents the incubation time before start of recrystallization, and they

re determined by a non-linear least square fitting procedure. The

emporal evolution of the recrystallized fraction determined from

he above model for different temperatures is shown in Fig. 3 b

solid lines). For illustration purpose, the data set of recrystallized

raction at 1350 °C, evaluated from hardness for different expo-

V. Shah, J.T.S. Beune, Y. Li et al. Journal of Nuclear Materials 545 (2021) 152748

Table 2

Temperature dependent time for half-recrystallization de-

termined using the JMAK model from Fig. 3 b.

T [ °C] t hal f

1450 5.3 min

1350 1.18 hr

1300 1.87 hr

1250 6.05 hr

1200 22.61 hr

s

t

(

F

d

w

m

0

w

t

t

[

t

p

r

f

t

t

c

r

w

d

t

o

e

p

o

(

t

p

F

t

s

[

f

o

(

n

t

r

o

e

5

m

v

b

g

m

H

a

T

(

W

e

s

N

t

i

[

t

c

d

e

a

t

p

t

n

t

[

H

i

d

a

k

t

t

e

a

p

ure times is also plotted in Fig. 3 b (circular markers), while the

ime for half recrystallization t hal f determined from these curves

Fig. 3 b) are provided in Table 2 . For the accounted temperatures in

ig. 3 b, the coefficients b and t inc tend to be strongly temperature-

ependent. At the same time, the Avrami exponent n displays a

eak temperature dependence, with an average value of 1.4 (for

ore details, refer to Appendix C ). This value is within the range

.78 to 2, reported by Alfonso et al. [ 34,35 ] and Wang et al. [36] for

arm-rolled tungsten with varying deformation levels. Fundamen-

ally, the characteristic time for the recrystallization process and

he temperature are known to follow an Arrhenius-type relation

5] , thereby allowing a measure of activation energy to be ob-

ained. This is demonstrated in Fig. 4 , where the aforementioned

arameter t hal f is plotted. For obtaining precise insights on the

ole of H atoms on the kinetics, the data set from the work of Al-

onso et al. [35] corresponding to oven annealed samples (1100 °C

o 1250 °C) is used for comparison, extrapolated linearly to higher

emperatures and is shown in Fig. 4 . The t hal f (blue square and cir-

ular markers) data points in Fig. 4 show similar behaviour as the

eference linear fit for the temperature range 1150 °C to 1350 °C

ith some scatter, while for higher temperatures, i.e. > 1400 °C, the

eviation from the reference tends to be higher. The error bars in

he t hal f data points account for the maximum temperature error

f the IR camera. The scatter of the t hal f could be due to the het-

rogeneous annealing approach used in the present work as com-

ared to oven annealing. Additionally, using the slope of a linear fit

f t hal f (not shown in Fig. 4 ), an activation energy of 433 kJ.mol -1

± 2%) is obtained. This activation energy is comparatively higher

han 352 kJ.mol -1 ( ± 4%) for a highly deformed warm-rolled W

late (90%), as reported by Alfonso et al. [35] . Moreover, Alfonso

ig. 4. Arrhenius plot depicting the time for half-hardness loss/half recrystalliza-

ion ( t hal f ) determined from indentation measurements following H plasma expo-

ure, in addition to data points from oven annealed (no H exposure) W samples

35] . The blue coloured square markers represent the time for half-hardness loss

rom Table 1 , while the circular markers represent the time for half-recrystallization

btained from Table 2 . The activation energy determined for H exposed samples

blue markers) is approximately 433 kJ.mol -1 , which is higher than for the oven an-

ealed non-exposed samples (approximately 352 kJ.mol -1 , red markers), but within

he range 377 to 460 kJ.mol -1 for grain boundary diffusion. (For interpretation of the

eferences to colour in this figure legend, the reader is referred to the web version

f this article.)

o

t

a

a

a

d

e

i

m

v

a

o

f

l

r

m

i

o

p

H

i

r

H

d

h

m

5

t al. [34] have also reported an activation energy of approximately

79 kJ.mol -1 for moderately deformed warm-rolled W plate (67%),

uch higher than the 90% rolled plate. The difference in the acti-

ation energies between the two plates has been further reasoned

ased on the microstructural features, specifically the low-angle

rain boundary spacing (LAGB), i.e. with increasing degree of defor-

ation (67% to 90%) the LAGB spacing reduces (2 μm to 0.67 μm).

owever, in the present case the relation between the lowering of

ctivation energy with decreasing LAGB spacing does not hold true.

he LAGB spacing in the present work is approximately 2.75 μm

see Appendix A ), marginally higher than the moderately deformed

plate. This discrepancy can possibly understood as due to differ-

nce between the forming process, i.e. bi-directional forging ver-

us rolling, and the consequent variance in the defect density.

onetheless, the determined activation energy tends to be within

he 377 to 460 kJ.mol -1 range for the grain boundary diffusion, and

s much lower than that of 502 to 586 kJ.mol -1 for bulk diffusion

37] . This therefore implies that the recrystallization kinetics tend

o proceed pre-dominantly by grain boundary diffusion.

Elaborating on the role of H for recrystallization kinetics, one

ould expect that retardation or acceleration of recrystallization

ue to H would dramatically influence the activation energy. For

xample, the diffusion of Ar atoms and presence of dissolved Ar

toms in W has been shown to delay the onset of recrystalliza-

ion, ultimately resulting in a higher activation energy as com-

ared to vacuum annealing [38] . Also, considering the case where

he H atoms form complexes and retard recrystallization by pin-

ing the grain boundaries (similar to second phase precipitates),

he activation energy would increase, as demonstrated by Zan et al.

39] for oxide reinforced W. On the other hand, for the case where

atoms tend to favour excess vacancy concentration by lower-

ng their formation energy [27,28] , the activation energy would

ecrease in comparison to the pristine state, as excess vacancies

ccelerate recrystallization [18,40] . However, following the current

inetic analysis and the observation of activation energy represen-

ative of grain boundary diffusion in the present work, identical

o the work of Alfonso et al. [35] , it can be concluded that low

nergy H plasma exposure of W has an inconsequential effect in

ccelerating or retarding the recrystallization process. One of the

lausible explanations for this behaviour can be the combination

f low implantation depth of H atoms and relatively high tempera-

ures during exposure, eventually resulting in high desorption rates

s well as limited long range bulk diffusion, and thus an over-

ll lower H bulk concentration. Additionally, since high temper-

tures prevent trapping events between the H atoms and lattice

efects (high energy traps such as dislocations have a de-trapping

nergy of 1.25 eV [41] ), pre-dominantly all the H atoms would ex-

sts as freely diffusing interstitial atoms, consequently contributing

arginally to elastic stresses in the lattice. In a nutshell, the obser-

ation of unaltered recrystallization kinetics despite long term heat

nd H plasma exposure could be due to the lower bulk retention

f H atoms and the resulting weak interaction with the lattice de-

ects. This conclusion is further supported by observations of neg-

igible deuterium (D) concentration in bulk W following nuclear

eaction analysis (NRA) of high flux, high fluence D exposure of W

onoblocks performed by Morgan et al. [42] . Nevertheless, further

nvestigations specifically addressing the oven annealing behaviour

f samples (like in the present work) will allow to decouple and

rovide additional insights into the effects of microstructure and

plasma. Moreover, distinguishing between the role of different

mpurities, i.e. H, He, and beryllium (Be), the major impact on the

ate of recrystallization is expected due to the He as compared to

, as it has a stronger binding tendency with lattice defects (pre-

ominantly promoting the formation of He bubbles). On the other

and, the role of Be is unclear, hence difficult to ascertain and de-

ands further investigation.

V. Shah, J.T.S. Beune, Y. Li et al. Journal of Nuclear Materials 545 (2021) 152748

4

c

i

h

t

t

i

b

e

c

s

m

d

o

a

D

u

D

c

i

C

i

F

-

i

M

M

t

t

l

a

A

k

F

N

R

t

r

u

p

C

p

o

s

A

(

c

d

F

fi

a

g

o

. Conclusion

Summarising, the present work dealt with investigating the re-

rystallization kinetics of W samples exposed to H plasma for vary-

ng time. The characteristic time of recrystallization process, i.e. for

alf-hardness loss or half-recrystallized fraction ( t hal f ) revealed ac-

ivation energy similar to that of the oven annealed samples in

he literature, with the recrystallization operating mechanism be-

ng grain boundary diffusion. This agreement on activation energy

etween the present work and literature indicates that under low

nergy H plasma exposure, the influence of H atoms on the re-

rystallization kinetics of W is not significant. Future developments

pecifically addressing the synergistic influence of bulk displace-

ent damage and H plasma exposure would allow a better un-

erstanding of the interplay between H atom and lattice defects

n the microstructure evolution, eventually allowing more accurate

nd realistic predictions for future fusion reactors.

ata availability

The raw or analysed data reported in this work are available

pon a reasonable request.

eclaration of Competing Interest

The authors declare that they have no known competing finan-

ial interests or personal relationships that could have appeared to

nfluence the work reported in this paper.

RediT authorship contribution statement

V. Shah: Conceptualization, Methodology, Visualization, Writ-

ng - original draft. J.T.S. Beune: Conceptualization, Data curation,

ig. 5. EBSD based characterization of microstructure in pre-exposed (deformed) state (a

gure (d) Orientation distribution function (in Euler space), with ϕ 2 sectioned at 45 °. Th

chord and the boundaries with misorientation less than 15 ° as shown in (b) . For statist

rain boundary spacing being approximately equal to 2.75 μm.(For interpretation of the r

f this article.)

6

ormal analysis, Investigation, Methodology, Visualization, Writing

review & editing. Y. Li: Conceptualization, Methodology, Visual-

zation, Writing - review & editing. Th. Loewenhoff: Investigation,

ethodology, Writing - review & editing. M. Wirtz: Investigation,

ethodology, Writing - review & editing. T.W. Morgan: Concep-

ualization, Funding acquisition, Methodology, Project administra-

ion, Supervision, Writing - review & editing. J.A.W. van Domme-

en: Conceptualization, Funding acquisition, Methodology, Project

dministration, Supervision, Writing - review & editing.

cknowledgements

DIFFER is part of the institutes organization of NWO. We ac-

nowledge the support of the Magnum-PSI Facility Team at DIF-

ER. The Magnum-PSI facility at DIFFER has been funded by the

etherlands Organisation for Scientific Research (NWO. ) and EU-

ATOM. This work has been carried out within the framework of

he EUROfusion Consortium and has received funding from the Eu-

atom research and training programme 2014–2018 and 2019–2020

nder grant agreement no. 633053 The views and opinions ex-

ressed herein do not necessarily reflect those of the European

ommission . Additionally, this research was carried out under

roject number T16010c in the framework of the Research Program

f the Materials innovation institute (M2i) ( http://www.m2i.nl )

upported by the Dutch government.

ppendix A. Microstructure characterization: pre-exposure

deformed) state

Fig. 5 schematically shows the EBSD characterization of the mi-

rostructure before H plasma exposure (deformed) state. The Y-

irection inverse pole figure (Y-IPF) map ( Fig. 5 a) clearly depicts

) Y-direction inverse pole figure (IPF) map (b) grain boundary map (c) Inverse pole

e low angle grain boundary spacing was determined from the intercepts between

ics, 50 chords were laid over the grain boundary map, with the average low angle

eferences to colour in this figure legend, the reader is referred to the web version

V. Shah, J.T.S. Beune, Y. Li et al. Journal of Nuclear Materials 545 (2021) 152748

Fig. 6. (a) Z-direction (depth) inverse pole figure (Z-IPF) map and the corresponding grain orientation spread (GOS) map (b) obtained from EBSD measurements following

3 hr H plasma exposure. The EBSD map shown in (a) and (b) corresponds to recrystallized regime indicated by the black box in 3 hr HV map ( Fig. 2 b). The recrystallization

extent along the plane of exposure is approximately 7 mm from the hot-spot centre.(For interpretation of the references to colour in this figure legend, the reader is referred

to the web version of this article.)

t

e

d

(

i

g

αt

{

A

e

l

(

r

T

s

i

b

o

r

A

t

s

e

J

t

t

Table 3

Temperature dependent JMAK model coefficients for recrystallization kinetics of de-

formed tungsten during H plasma exposure.

T [ °C] b n t inc (s)

1200 1.02 × 10 −5 1.3 500

1250 3.85 × 10 −5 1.2 138

1300 1.24 × 10 −4 1.5 0.76

1350 1.95 × 10 −4 1.4 1.2 × 10 −6

1450 2.54 × 10 −3 1.61 0.95

R

he bi-directional forging induced elongation of grains with an av-

rage aspect ratio of 4.5. The microstructure clearly shows a high

ensity of low angle grain boundaries (LAGB) due to deformation

Fig. 5 b), with an average LAGB spacing of 2.75 μm. The LAGB spac-

ng was calculated by counting the intercepts with LAGBs along a

iven line (shown in the magnified image Fig. 5 b). Pre-dominantly

fibre texture was observed ( Fig. 5 c) with relatively high intensi-

ies for { 001 }〈 110 〉 texture component, followed by { 112 }〈 110 〉 and

111 }〈 110 〉 texture components ( Fig. 5 d).

ppendix B. Microstructure characterization: H plasma 3 hr

xposure

The recrystallized region near the centre of the sample fol-

owing 3 hr H plasma exposure is shown in terms of Z-direction

depth) inverse pole figure (Z-IPF) map in Fig. 6 a. This region cor-

esponds to the black box indicated in the 3 hr HV map in Fig. 2 b.

o further ascertain the recrystallized state, the grain orientation

pread (GOS) map of the corresponding microstructure is shown

n Fig. 6 b, and as seen in the map, most of the grains tend to

e below the recrystallization threshold orientation spread value

f 3 °. Thereby, confirming the low HV values observed in the cor-

esponding region of the 3 hr exposure HV map ( Fig. 2 b).

ppendix C. Parametrization: JMAK model

Table 3 summarizes the temperature dependent coefficients of

he JMAK model determined using non-linear least square regres-

ion analysis. Note that only temperatures ≥ 1200 °C are consid-

red here. For lower temperatures, accurate parametrization of the

MAK model could not be obtained, as the kinetics of recrystalliza-

ion are slow, and hence experiments with longer exposure time

han considered in the present work are required.

7

eferences

[1] V. Philipps , Tungsten as material for plasma-facing components in fusion de- vices, J. Nucl. Mater. 415 (1) (2011) S2–S9 .

[2] R. Pitts , X. Bonnin , F. Escourbiac , H. Frerichs , J. Gunn , T. Hirai , A. Kukushkin ,E. Kaveeva , M. Miller , D. Moulton , V. Rozhansky , I. Senichenkov , E. Sytova ,

O. Schmitz , P. Stangeby , G. De Temmerman , I. Veselova , S. Wiesen , Physics ba-sis for the first ITER tungsten divertor, Nucl. Mater. Energy 20 (2019) 100696 .

[3] G. Federici , C. Skinner , J. Brooks , J. Coad , C. Grisolia , A. Haasz , A. Hassanein ,

V. Philipps , C. Pitcher , J. Roth , W. Wampler , D. Whyte , Plasma-material interac-tions in current tokamaks and their implications for next step fusion reactors,

Nucl. Fusion 41 (12) (2001) 1967 . [4] M. Merola , G. Vieider , M. Bet , I. Vastra , L. Briottet , P. Chappuis , K. Cheyne ,

G. Dell’Orco , D. Duglué, R. Duwe , S. Erskine , F. Escourbiac , M. Föbvre , M. Grat-tarola , F. Moreschi , A. Orsini , R. Pamato , L. Petrizzi , L. Plöchl , B. Riccardi ,

E. Rigal , M. Rödig , J. Salavy , B. Schedler , J. Schlosser , S. Tähtinen , R. Vesprini ,

E. Visca , C. Wu , European achievements for ITER high heat flux components, Fusion Eng. Des. 56 (2001) 173–178 .

[5] F. Humphreys , M. Hatherly , Recrystallization and Related Annealing Phenom- ena, Elsevier, 2012 .

[6] G. De Temmerman , T. Hirai , R. Pitts , The influence of plasma-surface inter- action on the performance of tungsten at the ITER divertor vertical targets,

Plasma Phys. Control. Fusion 60 (4) (2018) 044018 . [7] G. Pintsuk , I. Bobin-Vastra , S. Constans , P. Gavila , M. Rödig , B. Riccardi , Qual-

ification and post-mortem characterization of tungsten mock-ups exposed to

cyclic high heat flux loading, Fusion Eng. Des. 88 (9-10) (2013) 1858–1861 . [8] S. Panayotis , T. Hirai , V. Barabash , A. Durocher , F. Escourbiac , J. Linke ,

T. Loewenhoff, M. Merola , G. Pintsuk , I. Uytdenhouwen , M. Wirtz , Self-castel-lation of tungsten monoblock under high heat flux loading and impact of ma-

terial properties, Nucl. Mater. Energy 12 (2017) 200–204 .

V. Shah, J.T.S. Beune, Y. Li et al. Journal of Nuclear Materials 545 (2021) 152748

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[9] V. Shah , M. van Maris , J. van Dommelen , M. Geers , Experimental investigationof the microstructural changes of tungsten monoblocks exposed to pulsed high

heat loads, Nucl. Mater. Energy 22 (2020) 100716 . [10] Y. Li , T. Morgan , J. van Dommelen , S. Antusch , M. Rieth , J. Hoefnagels , D. Ter-

entyev , G. De Temmerman , K. Verbeken , M. Geers , Fracture behavior of tung-sten-based composites exposed to steady-state/transient hydrogen plasma,

Nucl. Fusion 60 (4) (2020) 046029 . [11] M. Wirtz , G. Cempura , J. Linke , G. Pintsuk , I. Uytdenhouwen , Thermal shock

response of deformed and recrystallised tungsten, Fusion Eng. Des. 88 (9-10)

(2013) 1768–1772 . 12] G. Pintsuk , T. Loewenhoff, Impact of microstructure on the plasma perfor-

mance of industrial and high-end tungsten grades, J. Nucl. Mater. 438 (2013) S945–S948 .

[13] C. Linsmeier , M. Rieth , J. Aktaa , T. Chikada , A. Hoffmann , J. Hoffmann ,A. Houben , H. Kurishita , X. Jin , A. Litnovsky , S. Matsuo , A. von Müller ,

V. Nikolic , T. Palacios , R. Pippan , D. Qu , J. Reiser , J. Riesch , T. Shikama ,

R. Stieglitz , T. Weber , S. Wurster , J.-H. You , Z. Zhou , Development of advancedhigh heat flux and plasma-facing materials, Nucl. Fusion 57 (9) (2017) 092007 .

[14] G. Davis , Embrittlement of tungsten wires by contaminants, Nature 181 (4617) (1958) 1198 .

[15] V. Shah , J. van Dommelen , E. Altstadt , A. Das , M. Geers , Brittle-ductile transi-tion temperature of recrystallized tungsten following exposure to fusion rele-

vant cyclic high heat load, J. Nucl. Mater. 541 (2020) 152416 .

[16] F. Haessner , H. Holzer , Boundary migration in neutron-irradiated copper bicrystals, Scr. Metall. 4 (3) (1970) 161–165 .

[17] F. Haessner , H. Holzer , The effect of fast neutron irradiation on the recrystal-lization of cold-rolled (110)[112] oriented copper crystals, Acta Metall. 22 (6)

(1974) 695–708 . [18] W. Vaidya , K. Ehrlich , Radiation-induced recrystallization, its cause and conse-

quences in heavy-ion irradiated 20% cold-drawn steels of type 1.4970, J. Nucl.

Mater. 113 (2-3) (1983) 149–162 . [19] M. Abd El Keriem , D. Van Der Werf , F. Pleiter , Helium-vacancy interaction in

tungsten, Phys. Rev. B 47 (22) (1993) 14771 . 20] K. Farrell , A. Schaffhauser , J. Houston , Effect of gas bubbles on recrystallization

of tungsten, Metall. Trans. 1 (10) (1970) 2899–2905 . 21] D. Snow , The recrystallization of commercially pure and doped tungsten wire

drawn to high strain, Metall. Trans. A 10 (7) (1979) 815–821 .

22] S. Nogami , S. Watanabe , J. Reiser , M. Rieth , S. Sickinger , A. Hasegawa , Improve-ment of impact properties of tungsten by potassium doping, Fusion Eng. Des.

140 (2019) 48–61 . 23] T. Miyazawa , T. Hwang , K. Tsuchida , T. Hattori , M. Fukuda , S. Nogami ,

A. Hasegawa , Effects of helium on mechanical properties of tungsten for fu- sion applications, Nucl. Mater. Energy 15 (2018) 154–157 .

24] W. Guo , L. Cheng , G. De Temmerman , Y. Yuan , G.-H. Lu , Retarded recrystalliza-

tion of helium-exposed tungsten, Nucl. Fusion 58 (10) (2018) 106011 . 25] K. Song , M. Thompson , G. De Temmerman , C. Corr , Temperature dependence

of retarded recrystallisation in helium plasma-exposed tungsten, Nucl. Fusion 59 (9) (2019) 096031 .

26] T. Loewenhoff, S. Bardin , H. Greuner , J. Linke , H. Maier , T. Morgan , G. Pintsuk ,R. Pitts , B. Riccardi , G. De Temmermann , Impact of combined transient

8

plasma/heat loads on tungsten performance below and above recrystallization temperature, Nucl. Fusion 55 (12) (2015) 123004 .

27] V. Goltsov , D. Glyakov , G. Zhirov , Influence of dissolved hydrogen on recrystal-lization and recovery of mechanical properties of deformed palladium, Int. J.

Hydrog. Energy 31 (2) (2006) 211–216 . 28] M. Martin , A. Pundt , R. Kirchheim , Hydrogen-induced accelerated grain growth

in vanadium, Acta Mater. 155 (2018) 262–267 . 29] V. Nikoli ́c , J. Riesch , R. Pippan , The effect of heat treatments on pure and

potassium doped drawn tungsten wires: part I-microstructural characteriza-

tion, Mater. Sci. Eng. A 737 (2018) 422–433 . 30] H. Van Eck , G. Akkermans , S. Alonso van der Westen , D. Aussems , M. van

Berkel , S. Brons , I. Classen , H. van der Meiden , T. Morgan , M. van de Pol ,J. Scholten , J. Vernimmen , E. Vos , M. de Baar , High-fluence and high-flux per-

formance characteristics of the superconducting magnum-psi linear plasma fa- cility, Fusion Eng. Des. 142 (2019) 26–32 .

31] G. De Temmerman , M. Van Den Berg , J. Scholten , A. Lof , H. Van Der Meiden ,

H. van Eck , T. Morgan , T. de Kruijf , P. Zeijlmans van Emmichoven , J. Zielinski ,High heat flux capabilities of the Magnum-PSI linear plasma device, Fusion

Eng. Des. 88 (6-8) (2013) 4 83–4 87 . 32] H. Van Eck , T. Abrams , M. Van Den Berg , S. Brons , G. Van Eden , M. Jaworski ,

R. Kaita , H. van der Meiden , T. Morgan , M. van de Pol , J. Scholten , P. Smeets ,G. De Temmerman , P. de Vries , P. Zeijlmans van Emmichoven , Operational

characteristics of the high flux plasma generator Magnum-PSI, Fusion Eng. Des.

89 (9-10) (2014) 2150–2154 . 33] M. Banisalman , S. Park , T. Oda , Evaluation of the threshold displacement en-

ergy in tungsten by molecular dynamics calculations, J. Nucl. Mater. 495 (2017) 277–284 .

34] A. Alfonso , D. Jensen , G.-N. Luo , W. Pantleon , Recrystallization kinetics of war-m-rolled tungsten in the temperature range 1150–1350 ◦C, J. Nucl. Mater. 455

(1-3) (2014) 591–594 .

35] A. Alfonso , D. Jensen , G.-N. Luo , W. Pantleon , Thermal stability of a highly-de-formed warm-rolled tungsten plate in the temperature range 1100–1250 ◦C, Fu-

sion Eng. Des. 98 (2015) 1924–1928 . 36] K. Wang , X. Zan , M. Yu , W. Pantleon , L. Luo , X. Zhu , P. Li , Y. Wu , Effects of

thickness reduction on recrystallization process of warm-rolled pure tungsten plates at 1350 ◦C, Fusion Eng. Des. 125 (2017) 521–525 .

37] E. Lassner , W.-D. Schubert , The Element Tungsten, Springer, 1999 .

38] Y. Aleksandrova , L. Aleksandrov , V. Eliséeva , Effect of inert gases on the recrys-tallization of tungsten, Sov. Mater. Sci. 2 (3) (1967) 234–237 .

39] X. Zan , M. Gu , K. Wang , L. Luo , X. Zhu , Y. Wu , Recrystallization kinetics of 50%hot-rolled 2% Y 2 0 3 dispersed tungsten, Fusion Eng. Des. 144 (2019) 1–5 .

40] M. Feller-Kniepmeier , J. Gobrecht , Influence of excess vacancies on recrystal- lization of high purity gold, Acta Metall. 19 (7) (1971) 569–576 .

41] Y. Li , T. Morgan , D. Terentyev , S. Ryelandt , A. Favache , S. Wang , M. Wirtz ,

J. Hoefnagels , J. van Dommelen , G. De Temmerman , K. Verbeken , M. Geers ,Three mechanisms of hydrogen-induced dislocation pinning in tungsten, Nucl.

Fusion 60 (8) (2020) 086015 . 42] T. Morgan , M. Balden , T. Schwarz-Selinger , Y. Li , T. Loewenhoff, M. Wirtz ,

S. Brezinsek , G. De Temmerman , ITER monoblock performance under lifetime loading conditions in Magnum-PSI, Phys. Scr. 2020 (T171) (2020) 014065 .