PISCES-B linear plasma facility at UCSD Plasma ......Plasma interactions with Be surfaces R. P....

U C S DUniversity of California San Diego

R� Doerner - 2nd Be CRP, August 18-19, 2014

Plasma interactions with Be surfaces

R. P. Doerner for members of the PISCES Team

Center for Energy Research, University of California – San Diego, USA

Work performed as part of: D�� - Knoxville)

US-EU Collaboration on Mixed-Material PMI Effects for ITER

US- Japan Technology Exchange Program

PISCES-D3D Collaborations U C S DUniversity of California San Diego


PISCES-B linear plasma facility at UCSD PISCES



PISCES I��

Ion flux (cm2s–1

) 1017–10

19 ~10

19 - 10

20

Ion energy (eV) 20–300 (bias) 10–300 (thermal)

Te (eV) 4–4� 1–1��

ne (cm–3

) 1012–10

13 ~10

13

Be Imp. fraction (%) Up to a few % 1–10 (ITER)

Pulse length (s) Steady state 1000

P�I � !�r" #$ C% &% '� C% &% '� ((P# $� $)�*"�$ H% +% H� H% +% �% H�

PISCES-B experimental arrangement

PISCES



Outline of Presentation

• Erosion of Be in a high-flux plasma environment

– Erosion induced morphology

– Sticking coefficient of Be (cavity probe measurements)

– Gross vs. net erosion temperature dependence

• Al erosion behavior (surrogate for Be)

- Al retention measurements

• Be on W studies

- Impact of transients

• Discussion

PISCES

B, -./023,

before plasma

exposure

after

Development of surface morphology decreases erosion

s5��6s�6578

mass loss:

Þ morphology change can account for a

factor of 2 in reduction of the yield

PISCES

9 1 2 3 4 5

0

2

4

6

8

10

12

:;<==>?@A>EFGJK-3

Be p

er

ion

)

ion fluence (1022

/(cm2s))

D2, <330K, -100V

SLMNNOQUVW XUOYZ [\ ]O ZO^QO_`O` ZMQUVW LMQO a

plasma exposure, penetration distance of Be into

plasma increases with fluence

bcbd

0.1

0 10 20 30 40 50

fg hi j kgl mno pqtmuvqw02 (t = 128 s): fit (z = 0-10 mm)15 (t = 4746 s): measured15 (t = 4746 s): fit (z = 0-10 mm)

z xyy{

|e ~ 20.7 mm

le ~ 28.3 mm

}~�� x�i ~ 100 eV)

PISCES

�

0.04

0.08

0.12

0 1000 2000 3000 4000 5000 6000

��

��

D on Be

(Ei ~ 100 eV)

�� ¡�¢£¡¤¡¥¦ §�¨�¤¡¢© ��¡ ��¡©ª¡«¬ ©��© ®ª£ fluence.

Characteristic length changes with incident ion energy.

¯° ±° ²°

PISCES

³´µ ¶·¸¹º» ¼½ ¾º ¿¸À¹º¿» ¸½¿ºÀ ÁºÂÃ ·-2 deuterium fluence at;

a) -170 Vbias, b) -100 Vbias and c) -50 Vbias.

ÄÅ ±° ²°

SEM images of Be targets at -100 V bias and deuterium fluence at;

a) 0 fluence, b) 3e25 m-2 and c) 3e26 m-2 fluence.


ÆÇ Doerner - 2nd Be CRP, August 18-19, 2014

D (& He) plasma erosion rates consistently lower than

TRIM, Ar sputter data is higher than TRIM

ÈÉ-5

10-4

10-3

10-2

10-1

0 20 40 60 80 100 120 140

ÊËÌÍÎÏËÐ ÑÒËÓÐ ÒÔ Õ ÖÓÌÍ×ÌY_D(Ei)2*Y_D(Ei/2)3*Y_D(Ei/3)total_Y(0.25,0.47,0.28)

Øi [eV]

Ù ÚÛ ÜÝ

[from Doerner et al., PSI20]

PISCES

ÞßÞÞÞà

0.001

0.01

0.1

1

0 50 100 150 200 250

Ar on Be

Yexp corrected for 25%++TRIM 25%++TRIM singly ionized

áâãä åæçèãéê ëìíî

ïZZUNU[V [\ ðñò Ar to D plasma is sufficient to

prevent surface morphology and prevent decrease in

sputtering yield during high fluence exposure

ó

0.1

0.2

0.3

0.4

0 500 1000 1500 2000 2500 3000 3500

ôõö÷ øùú

ûü÷ýþÿ÷A � � ��öö

D+Ar(~10%) on Be

(Ei ~ 100 eV)

0�0�

0.06

0.08

0.1

0.3

0.5

0 10 20 30 40 50

�� 03 (t = 96 s): fit (z = 0-10 mm)25 (t = 3431 s): measured25 (t = 3431 s): fit (z = 0-10 mm)

z ��

le ~ 15.3 mm

le ~ 16.2 mm

D�� !" #$ �%i ~ 100 eV)

PISCES



Distinct surface morphology differences exist

after high and low mass sputtering of a Be target

• Smooth Be surface exists after Ar plasma sputtering

• No significant decrease in Be erosion with fluence during Ar

sputtering

After Ar plasma sputtering (~75 eV) After D plasma sputtering (~80 eV)

PISCES

EYU&UV_NU[V [\ `'_YY[( _VWYO `LMNNOQUVW )ZMO N[

morphology increasing redeposition) leads to an

increase in penetration distance PISCES

*+,

1

10

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6

l simulation

= 1.5 cm

F-./ 1234.56 7819/39:85 ;81/38<2/89=>Surface morphology (45 degree angular cutoff)

Z ?@BC

l = 0.75 cm

l = 1.32 cm

[GHÚI Nishijima et al., JNM 415 (2011) S96]

J = 15 mm R plasma = 30 mm

LKÛÝM ÚG MKNOP

Z



Higher surface temperature decreases morphology

SEM images of beryllium surfaces exposed to deuterium plasma for

one hour, and -60V, bias at a) 300 K, b) 410 K, c) 570 K,d) 770 K and e) 870 K.

PISCES

e)



Comparison to Yamamura’s [2] model of optimum sputtering angles

aQ bQ

D

Plasma

-115 V

300 K

He

Plasma

-115 V

300 K

PISCES

STU VÇVWXWXY\W] VÇ Itikawa, and N. Itoh, “Angular dependence of sputtering yields

of monatomic solids,” Report No. IPPJ-AM-26, 1983.

^_`_`cd_fg hid`cj_ kdmnopqg gr_dkmd pismg hid tu pi`k_dmn qi vm

and significantly less sharp cones for Ar bombardment, in agreement

with experimental observations.



Should morphology develop on ITER’s Be plasma-facing surfaces?

• Present-day confinement device do not usually observe surface morphology on post-mortem inspection of plasma-exposed surfaces

• Present confinement devices employ various/changing experimental geometries which is likely to inhibit morphology development

• Flattop operation is comparable to plasma ramp up and ramp down phases in today’s devices

• Existing devices usually employ inertial cooling, which means elevated temperatures are encountered, decreasing development of morphology

• ITER will have active cooling and relatively long pulses, so morphology is more likely to develop

• JET-ILW Be samples need to be examined for evidence of morphology

PISCES

wxyx{| }~��-21) result from JET-ILW do hint at morphology developing

on plasma-exposed Be surfaces

PISCES

�� Bergsaker et al., P02-093, submitted to JNM under review]

bQ

�Q

JET-ILW PISCES-B

Top

view

30° tilt

of same

surface





� Erosion induced morphology

� Sticking coefficient of Be (cavity probe measurements)

� Gross vs. net erosion temperature dependence



• Be on W studies


• Discussion

PISCES



Recall: Net erosion does not change as expected

• Use reproducible plasmas while

simply changing the Be oven

temperature

• Net erosion remains ~ constant

until Be influx >> sputtering rate

• Seeding rate must exceed erosion

rate by a factor of 10 to reach net

deposition

• Two possibilities, reduced

sticking of depositing Be, or

increased re-erosion, could

explain observations

D/He/Be plasma,

50 eV ion energy

Weight loss measures net erosion

��

0

0.0005

0.001

0.0015

0.002

0.0025

10-5

0.0001 0.001 0.01 0.1

Experiment

Prediction

��

Net deposition

PISCES


�� Doerner - 2nd Be CRP, August 18-19, 2014

�� ¡¢£ ¤¥¦§¨© ª�« §¨ ¬©¨ ¢¦ ®¨�©¬¥¨ ©¢¡ª¯¡«° ª¦¨±±¡ª¡¨«¢

Use He plasma to reduce

uncertainties of BeD

PISCES



Results from cavity probe diagnostic

²³´�eµ¶·a¸ ¹eº·¸¶

Experimental realization

PISCES



Cavity probes show significant sticking of seeded Be ions PISCES

»¼½¾¼¿À½Á ÂÃÄÅÁÄ ¿¼Å½Ã¾Å Æ¼ÇÈÁ¼ ÁÈ¿¼ ÇÈÁÂÃÀÇÄ¾¼É PISCES



Net vs. gross erosion revisited

Ê·¹Ëa�e ¸³ºº µ¹³babÌ¸Ì¶Í be¶Îee´ ÏÐÑ a´Ò Ó �a´´³¶ explain Be seeding experimental results

PISCES





� Erosion induced morphology


� Gross vs. net erosion temperature dependence

Ô Al erosion behavior (surrogate for Be)


• Be on W studies


• Discussion

PISCES



New net erosion measurements (while varying only cooling rate)

confirm previous temperature dependent erosion rates

ÕÖÕÕÕ×

0.001

0.01

400 500 600 700 800 900 1000 1100 1200

Mass loss (@ 100 eV)Normalized Be I spectroscopyNormalized Be I spectroscopy [from 3]

ØÙÚÛÜÝÞ ßÞàáÞÚÜßÙÚÞ âãä

PISCES

�� å. P. Doerner

et al., J. Nucl. Mater.

257 (1998) 51]

[3] R. P. Doerner et al.,

J. Nucl. Mater.

337-339 (2005) 877

Recall the development of surface

morphology is temperature dependent

and morphology reduces erosion rate.

But net erosion is temperature

independent (until very high temperature).

Possible explanation could be the

temperature dependent erosion rate of

BeD (decreases with increasing temperature)

compensates for changes in erosion

rate due to morphology (increases

erosion as temperature increases).



Expected relationship between net and gross erosion is recovered at elevated temperature {N=G*(1-R)}

æçèçççé

0

0.0005

0.001

0.0015

0.002

0.0025

0.003

10-5

0.0001 0.001 0.01 0.1

weight loss w seeding

Net yield (@~900K)Net prediction (high T)Net yield (@370-500K)êëì íîëïðñìðòó ôõòö ÷ø

ùëîúõõðûü ðüíûîðìú ñòóñëóìîýìðòó

Net deposition

PISCES

þÿR��

Two possibilities, reduced sticking

of depositing Be, or increased

re-erosion, could explain observations.

Depositing Be may behave differently

from lattice Be atoms.

Enlist MD calculations to help

understand deposited atom behavior

at different surface temperatures.



M ��

surface atoms created during stopping of incoming ions

• A Be 0001 surface, containing a

raised island, is created and a

Be adatom is placed on the top

surface of the island at

different surface temperatures

• The adatom is followed until it

descends the step at the edge

of the island, or until the total

time of the simulation exceeds

1 ns

PISCES



PISCES

M ��

• Below 500 K, the adatom is never observed to descend the step edge (within the time of the MD simulation, ~ 1ns)

• Between 500 and 1000 K a clear Arrhenius relationship (with an activation energy of 0.32 eV) is seen in the rate at which the adatom descends the step

• A nudged-elastic band (NEB) [6] method was also used to calculate the migration energy and Ehrlich-Schwoebel barrier at the edge of the island and obtained 0.22 eV for the migration energy and 0.33 eV for the Ehrlich-Schwoebel energy barrier



MD simulation results interpretation

• Surface atoms diffusing toward a

terrace edge from above and below

the ledge experience different forces.

The so-called Ehrlich-Schwoebel

energy barrier tends to reflect atoms

approaching to descend the step,

whereas atoms attempting to ascend

tend to be absorbed by the edge,

causing the layer to grow

• An atom absorbed at the end of the layer will be bound more tightly than a

mobile surface atom. Coupling the slower surface diffusion at lower temperature

with reflections from the Ehrlich-Schwoebel barrier can result in longer lifetimes

of less tightly bound surface atoms.

• The longer lifetime could result in an increased re-erosion probability in the case

of Be-seeded atoms at low surface temperature and be responsible for the added

deposition needed to achieve no mass loss from the plasma exposed surface

Ediff E

EE-S SSS

PISCES




� Erosion of Be in a high-flux plasma environment




� Al erosion behavior (surrogate for Be)


� Be on W studies


• Discussion

PISCES



Measurements of Aluminum Erosion and

Deposition in DIII-D and Pisces-B

July 14, 2014 Presented by: Chris Chrobak P.C. Stangeby, A.W. Leonard, D.L. Rudakov, C.P.C. Wong, A.G. McLean, G.M. Wright, D.A.

Buchenauer, J.G. Watkins, W.R. Wampler, J.D. Elder, R.P. Doerner,

D. Nishijima, and G.R. Tynan

1. Aluminum Erosion/Deposition Experiments in DIII-D

1. Direct measurement of erosion and deposition (ex-situ)

2. Spectroscopic measurement of erosion (in-situ)

2. Aluminum Erosion Experiments in PISCES-B

1. Determine the inverse photon efficiency (S/XB) factor for Al

2. Compare results from He and D plasmas

3. PISCES-B Experiments Provide Needed S/XB Data



Redeposited Al Measured Using Rutherford Backscattering Before and After Exposure

� Gross Erosion: >6.7nm/s

• Net Erosion: 5.6 nm/s

• No detectable re-deposition

• G�� !�"# $%& "'(�

• N)* ��!�"# +%& "'(�

• B,�)-. /)*)0*,1-) �)-deposition

234 -10 0 10 200

15

150

300

450

600 567896:7;69

Al C

overa

ge

(10

15 a

tom

s/c

m2)

Toroidal Distance from Center (mm)<=> -10 0 10 20

0

15

150

300

450

600 ?@ACD@After

Al C

overa

ge

(10

15 a

tom

s/c

m2)

Toroidal Distance from Center (mm)

1/e =

5.9mm

BT, Flow BT, Flow



Spectroscopic Measurement of Erosion

FFH IJKLO PQ STUIV WJXYYXPZ [X\]

minor background emission from D & C

Emission Spectrum and

CCD spectral filter

^_`a-Averaged Photon Flux:

Al Atomic Flux Using

Transmission-Averaged S/XB



• Post-Mortem Nuclear Reaction Analysis of DIII-D exposed Al Samples:

– Erosion Yield ~ expected D→Al sputtering yield

– Net Erosion ~ 70% Gross Erosion (low local redeposition)

• Spectroscopic Measurements of Al(I) Erosion Found:

– Large disagreement and with NRA-observed erosion rate

Initial Results from DIII-D and PISCES Al Erosion Experiments

0.0

0.5

1.0

1.5

2.0

2.5

bcd

Gro

ss

Sp

ec

e Al Ero

sio

n R

ate

(n

m/s

)

fg 20

30

50

100

200

300

500

1000

2000

3000

5000

10

100

1000

10-3

10-2

10-1

hijklmnlopnqirstulvmwnlox

Incident D Energy (eV)

yzzz{y |} ~�� }��z��~�{� y{�|}

��} z�� y{�|} (Expt)

Normal Inc D->Al (Model)

�� {� |} (Model)



Measurement S/XB in PISCES He plasmas

• �-4x higher than computed ADAS values, but still 10x lower than

needed for DIII-D Yield Measurements to Agree w/ Spectroscopic

Yield

• �� ¡��¢£¤¥ ¦§� �¡ �¥��¨��¢� �©�ª�� ¢ª � §��¢ª�yield during run (surface roughening)

« 10 20 300.1

1

10

¬®¬ ¯¬°±¬±² ³´«

µ²¶·²®¸ ¯¹º°±¬±² ³´

ADAS calculated

Al-I 396.1

nm

S/X

B

Te (eV)



S/XB in D-plasmas consistently larger than in He-plasmas

• Suggests D-Al interaction leads to additional Al

erosion

– Chemically-based mechanism?

» 10 20 300.1

1

10

¼½½½¾¼ ¿¼À

Á¼ÁÂ ÃÄ»

Å½ÂÆÇÂ¾È ¿¼ÀÅ½ÂÆÇÂ¾È ¿ÉÊÀ

Á¼ÁÂ ÃÄ

ADAS calculated

Al-I

396

.1n

m S

/XB

Te (eV)


ËÌ Doerner - 2nd Be CRP, August 18-19, 2014

• PISCES Experiments Show Al-D emission spectrum originating from the

sample

• Spectrum & Al-H band structure (scaled to Al-D center peak) matches

published locations

• Chemical erosion increases wt. loss but not Al (I) emissivity è Impacts

S/XB values

Evidence for Chemical Erosion of Al in D Plasmas

ÍÎÎ 424 426 4280

1000

2000

3000

4000

ÏÐÑÒ ÓÔÕÖ ÏÑ×Ø ÙÚÚ

Al IIAl-D ?

ÛÜÝÞßÞàáâã äàåæ

çèéêëìêí îïðééðñò óôêõöë÷ïø÷èùðéúêí ûùüý þðòêþðòê ÿíêòöð ðõ�öðñò��òí óöë÷õö÷ëê �óõ� ùêí ëñï ûùü��

422.

67

423.

57

424.

83


ËÌ Doerner - 2nd Be CRP, August 18-19, 2014 U C S DUniversity of California San Diego




P��

Al samples, similar technique was used with Be samples to

minimize loss fraction of sputtered particles from the plasma column

PISCES



A �� A�A�

PISCES




� Erosion of Be in a high-flux plasma environment






� Be on W studies


� Discussion

PISCES



Laser on Be-coated W creates rich variety of

morphology changes

7� ��2/s1/2

Ablation of Be, recrystallization of W

Be-W alloying and cracking

Delamination and

holes in Be coating

Be melt flow

D !"#$%#& '( %)*+,- ./ "#0/+

41

V2356

153 mm

Large type I ELM

PISCES



S89::;< =>S ?;8:@ 8<BCEF;G G@BH I;-W mixture at

center of spot

No laser

Damaged Be area

J 2 1

1

2

K

T. Höschen

Center of spot

70 MJ/m2/s1/2

LMNOQ Rmm) 4T

W

Be

PISCES



Sacrificial Be layer

UWX YZ[\Y\ ]W^^ _W`YZaW^ `bW` ac \Yd[ca`Y\ [X `bY divertor provide protection

to the underlying W? How thick does Be layer need to be?

Energy density required to sublimate (or equivalently, melt and evaporate) a

layer of Be with thickness d :

r is density, m is molar mass, and enthalpy of formation for Be is DH = 324 kJ/mol (CRC Handbook)

mre /dHBe

D=

How much energy is needed to remove Be coating?

efghi ihjklhn ophq rfshienergy density > eBe

Be layer survives

and protects W

tuvw uxyz t v{|}

44

�~� ��

��e

PISCES


ËÌ Doerner - 2nd Be CRP, August 18-19, 2014 4�

�-9 9-7 9-3

�� -2, 119 MJm-2s-1/2, 10 ms �� -2, 75 MJm-2s-1/2, 3 ms �� -2, 42 MJm-2s-1/2, 1 ms

9-9

R. Doerner 2nd Be CRP, August 18 19, 2014

SEM images support simple estimate for

Be layer removal

�� ¡¢�£ ¤¥ ¦§��when the absorbed laser

energy is greater than

the enthalpy of

formation of Be layer,

elaser > eBe

Laser spot

¨ Be

PISCES

©ª««¬®¯°±²³ª²²±´µ

• Knowing the state of the surface during the plasma interaction is

critical to developing an understanding of PMI

– Quantify the amount of imbedded plasma atoms in the surface

– Plasma quickly cleans impurities from the surface at the beginning of

exposures, but the plasma contains impurities that can effect PMI

– Mobility of surface atoms is likely important

• Close to unity sticking coefficient of Be has been measured

• Al seems to be a good erosion proxy for Be, but retention differs

drastically

• High-flux of low-z ions still appears to reduce erosion of surfaces

(JET-ILW sees lower Be erosion than PISCES-B and also seems to see

similar types of morphology developing on plasma-exposed Be)

• However, why does a high-flux of high-z ions appear to not affect

erosion?

PISCES

PISCES-B linear plasma facility at UCSD Plasma ......Plasma interactions with Be surfaces R. P....

Documents

Transcript of PISCES-B linear plasma facility at UCSD Plasma ......Plasma interactions with Be surfaces R. P....