U C S DU niversity o f C alifo rn ia S a n D iego
R. Doerner, IAEA CRP on H in Materials, Vienna, Sept. 26, 2006
Mixed-material studies in PISCES-B
R. P. Doerner, M. J. Baldwin, J. Hanna and D. Nishijima
Center for Energy Research, University of California – San Diego
R. Pungo, K. Schmid, and J. Roth
Max-Plank Institute for Plasmaphysics, Garching, Germany
Work performed as part of US-EU Collaboration on Mixed-Material PMI Effects for ITER
U C S DU niversity o f C alifo rn ia S a n D iego
R. Doerner, IAEA CRP on H in Materials, Vienna, Sept. 26, 2006
Outline
• Introduction
• Technical results– Temporal behavior of chemical erosion suppression– Response of Be/C to thermal transients– Be/W formation conditions
• Summary of possible mixed-material implications for ITER
U C S DU niversity o f C alifo rn ia S a n D iego
R. Doerner, IAEA CRP on H in Materials, Vienna, Sept. 26, 2006
PISCES-B has been modified to allow exposure of samples to Be seeded plasma
Radial transport guard
102 mm
153 mm
76 mm
195 mm
45 o
Cooled target holder
Heatable deposition probe assembly
Thermocouple
Thermocouple
Water cooled Mo heat dump
Resistive heating coils
High temperature MBE effusion cell used to seed plasma with evaporated Be
12 °
PISCES-B PlasmaTarget
Depositionprobesample
Axial spectroscopic field of view
Berylliumimpurityseeding
Radial transport guard
102 mm
153 mm
76 mm
195 mm
45 o
Cooled target holder
Heatable deposition probe assembly
Thermocouple
Thermocouple
Water cooled Mo heat dump
Resistive heating coils
High temperature MBE effusion cell used to seed plasma with evaporated Be
12 °
PISCES-B PlasmaTarget
Depositionprobesample
Axial spectroscopic field of view
Berylliumimpurityseeding
P-B experiments simulateBe erosion from ITER wall,subsequent sol transport and interaction with W bafflesor C dump plates, as well asinvestigation of codepositedmaterials using witness plates
U C S DU niversity o f C alifo rn ia S a n D iego
R. Doerner, IAEA CRP on H in Materials, Vienna, Sept. 26, 2006
A small beryllium impurity concentration in the plasma drastically suppresses carbon erosion
-50 V bias, 200ºC, Te = 8 eV, ne = 3 e 12 cm-3
Chemical erosion Physical sputtering
500
1000
1500
2000
2500
3000
3500
4000
No Be injection0.2% Be ion concentration
Wavelength (nm)
CD band
D gamma Be I
459445431 452438
4000
8000
1.2 104
1.6 104
2 104
2.4 104
2.8 104
No Be injection0.2% Be ion concentration
Wavelength (nm)
C I
941.5940.5939.5938.5937.5
U C S DU niversity o f C alifo rn ia S a n D iego
R. Doerner, IAEA CRP on H in Materials, Vienna, Sept. 26, 2006
Erosion suppression exhibits a temporal evolution (Be/C)
• Understanding the temporal behavior is critical to determining the fundamental mechanisms responsible for erosion mitigation
• PMI modeling codes should be able to reproduce temporal behavior to provide confidence
10-4
10-3
10-2
10-1
100
101
0 100 200 300 400 500 600
Inte
nsity
[a.u
.]
Time [s]
ID
IBeI/ID
ICD/ID(near)-ICD/ID(far)
20060322
Be/C = 86±1 sec
U C S DU niversity o f C alifo rn ia S a n D iego
R. Doerner, IAEA CRP on H in Materials, Vienna, Sept. 26, 2006
XPS data shows Be2C formation in resultant mixed-material surface
• Virtually all C remaining at the surface is bound as carbide (t > Be/C)
• Presence of carbide inhibits chemical erosion of C
• Carbide layer reduces sputtering yield of bound Be
• Subsequently deposited Be can be more easily eroded
• Codeposits are primarily Be once carbide layer forms
XPS analysis of Be on C sample surface[M. Baldwin et al., in press JNM]
U C S DU niversity o f C alifo rn ia S a n D iego
R. Doerner, IAEA CRP on H in Materials, Vienna, Sept. 26, 2006
If Be acts like B doping, then each Be atom should inhibit two C atoms from chemically eroding
• Chemical erosion model [Schenk
et al. JNM 220-222(1995)767] predicts boron reduces sp2 component in favor of sp3 hybridization. In other words, each B atoms inhibits a C=C bond, thereby affecting 2 C atoms
• In-situ Be seeding data shows similar behavior of chemical erosion mitigation, i.e. each Be surface atom is consistent with inhibiting 2 C atoms from chemically eroding
0
0.2
0.4
0.6
0.8
1
0 0.2 0.4 0.6 0.8 1
Y chem vs Be surface fraction
PISCES-B dataB doping modelCoster's PSI equation
Be surface fraction
U C S DU niversity o f C alifo rn ia S a n D iego
R. Doerner, IAEA CRP on H in Materials, Vienna, Sept. 26, 2006
In-situ Be doping of graphite exhibits similar behavior to boron doping of graphite
• Dopant increases retention
• Dopant shifts hydrogenic release to lower temperature
Temp (K)
Time (sec)
From A. Schenk JNM 220-222(1995)767 (B doping).Solid line Be seedingDashed line no Be seeding
U C S DU niversity o f C alifo rn ia S a n D iego
R. Doerner, IAEA CRP on H in Materials, Vienna, Sept. 26, 2006
Increased retention due to dopant becomes less significant during higher temperature exposures
Ts (K)
500 1000
Ret
entio
n (D
ato
ms
/m2 )
1020
1021
1022
1023
C target
C target, fBe~0.001
Be target, fCD4~0.02
Be target
Doerner et al. (1999)
Present study
• Solid symbols indicate mixed Be/C surfaces formed during plasma exposure, open symbols are from clean substrate data
• Mixed Be/C layers retain more deuterium than either Be samples or carbon samples exposed to similar plasma
• Difference in retention becomes less pronounced during higher temperature exposure
• Plasma contacting surfaces are not ITER’s main tritium inventory concern
U C S DU niversity o f C alifo rn ia S a n D iego
R. Doerner, IAEA CRP on H in Materials, Vienna, Sept. 26, 2006
Be first shuts down C chemical erosion, then subsequent Be re-erodes from surface
• Evolution of a mixed Be/C surface– Be oven opens at t = 0 sec.– Be ions arriving at t < 50s
shut down chemical erosion by forming Be2C surface layer [Baldwin JNM 2006 available on-line]
– Once Be2C is formed, subsequent Be arriving (T > 50 s) is more easily eroded and begins coating windows
– Be2C surface thickness saturates after carbide forms 50s in this exposure [Baldwin JNM 2006]
– Resultant codeposited material is primarily Be [Baldwin JNM 337-339(2005)590]
0
0.5
1
1.5
2
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0 500 1000 1500 2000
20060718_all [BeII/ne ~ 0.06%]
BeI/Dg near
Window transmission
CD/Dg near
Time [s]
0
0.5
1
1.5
2
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0 100 200 300 400 500
20060718_all [BeII/ne ~ 0.06%]
BeI/Dg near
Window transmission
CD/Dg near
Time [s]
U C S DU niversity o f C alifo rn ia S a n D iego
R. Doerner, IAEA CRP on H in Materials, Vienna, Sept. 26, 2006
Be2C layer thickness saturates
• Be layer observed after ~3E22 Be+/m2 (i.e 1600 s) accounts for virtually all incident Be
• Be layer after 1E23 Be+/m2 (i.e 4800 s) accounts for only ~30% of incident Be
• Be/C under these plasma exposure conditions would be ~ 2000 sec
Scattered energy (keV)
0.5 1.0
Cou
nts
(A
rb. u
nits
)
01
C
~ Depth (m)
0.00.5
x103
Be+fluence ~
3 x 1022 m-2
STD
Be+fluence ~
1 x 1023 m-2
O
Be
From M. Baldwin et al., in press JNM
RBS spectra of P-B samples exposed to unseeded and Be-seeded plasma
U C S DU niversity o f C alifo rn ia S a n D iego
R. Doerner, IAEA CRP on H in Materials, Vienna, Sept. 26, 2006
WPM samples show collection of beryllium-rich codeposits during Be seeding runs
Carbon target : 300ºC target exposure Carbon target : 700ºC target exposure
0 5000 10000 15000 20000 25000 30000 35000 40000 45000
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
100500
com
po
sitio
n (
%)
Depth (nm)
Be% C% O% Ta%
0 20 40 60 80 100 120
0
10
20
30
40
50
60
70
80
90
100
100500
Be% C% O% Ta-%
Ato
mic
%
Depth (nm)
More C is detected in codeposits during lower C target temperature exposure (possibly due to a combination of lower chemical erosion yield and/or quicker beryllium carbide layer formation)
U C S DU niversity o f C alifo rn ia S a n D iego
R. Doerner, IAEA CRP on H in Materials, Vienna, Sept. 26, 2006
Chemical erosion suppression time (Be/C) depends on several variables that can be varied almost independently
101
102
103
104
105
10-4 10-3 10-2 10-1
i ~ 1e22 m-2s-1, Ts ~ 690±10 K
i ~ 3e22 m-2s-1, Ts ~ 600±50 K, Ref. [2]
i ~ 3e22 m-2s-1, Ts ~ 955±20 K
Be/C
[s]
cBe
= 1.3e-3 = -2.1
= 1.1e-4 = -2.0
= 5.4e-4 = -2.1
Be/C = cBe
Ei ~ 40 eV 101
102
103
101 102
Be
/C [
s]
Ei [eV]
cBe = 10-3, = -2
Ts ~ 852-973 K, i ~ 1.7-4.5x1022 m-2s-1
Be/C = 1.5xEi1.2
102
103
0.001 0.0012 0.0014 0.0016 0.0018
Be/
C [
s]
1/Ts [K-1]
Ei ~ 30-40 eV, i ~ 1-3x1022 m-2s-1
cBe = 10-3, = -2
Be/C = 1.0 exp(4.6x103/Ts)
Be concentrationin plasma
Incident ionenergy
Surface temperatureof target From D. Nishijima et al., PSI17.
U C S DU niversity o f C alifo rn ia S a n D iego
R. Doerner, IAEA CRP on H in Materials, Vienna, Sept. 26, 2006
PISCES chemical erosion mitigation time scaling predicts suppression between ELMs in ITER
From D. Nishijima et al., PSI17. • Surface temperature effects reaction rate
• Be plasma concentration effects arrival rate at surface
• Ion energy effects erosion rate
• Ion flux impacts through redeposition
• Type of graphite does not seem to play a significant role (ATJ vs. CFC)
• Scaling law using these variables has been developed to allow extrapolation to ITER
conditions (Be/CITER
~ 6 msec) [cBe = 0.05, Ei = 20 eV, Ts = 1200 K and i = 1023 m-2s-1]
100
101
102
103
104
105
100 101 102 103 104 105
Be
/Ce
xp [
s]
Be/Cscale [s]
Closed: Ts ~ 550-700 KOpen: Ts ~ 800-970 K
Ei ~ 15 eVEi ~ 30-40 eVEi ~ 85 eV
3e 4 cBe 1e 2
15Ei [eV]85
550Ts [K]970
1i [e22 m 2s 1]4.5
Be/Cscale [s] = 1.0x10-7 cBe
-1.9±0.1 Ei0.9±0.3 i
-0.6±0.3 exp((4.8±0.5)x103/Ts)
x
X =CFC
U C S DU niversity o f C alifo rn ia S a n D iego
R. Doerner, IAEA CRP on H in Materials, Vienna, Sept. 26, 2006
Thermal transient experiments:Motivation for positive pulse biasing
• PISCES has shown that Be plasma impurities suppress carbon target erosion at temperatures up to 1000°C
• ITER will experience large temperature excursions (up to 3800°C) at the carbon dump plates during periodic ELMs
• Will the thin, surface Be, Be/C layers survive such dramatic temperature excursions?
• How will Be-W react during temperature excursions?
• It is possible to simulate the large temperature excursions associated with ITER ELMs in PISCES-B using positive sample biasing during plasma discharges.
U C S DU niversity o f C alifo rn ia S a n D iego
R. Doerner, IAEA CRP on H in Materials, Vienna, Sept. 26, 2006
Large power loads can be drawn to P-B sample during positive biasing
• During 1.5 MW/m2 power pulse graphite surface temperature rises to ~2000°C (by pyrometers)
• Bulk graphite temperature rise at back of sample ~20°C during 0.1 s. pulse (thermocouple)
• Surface temperature rise is limited by power supplies (IPP has supplied a new power supply as part of US-EU collaboration)
U C S DU niversity o f C alifo rn ia S a n D iego
R. Doerner, IAEA CRP on H in Materials, Vienna, Sept. 26, 2006
10-3
10-2
10-1
100
0 100 200 300 400 500 600
w/o H.P. (20060322)exponential fit for w/o H.P.w/ H.P. (20060323)exponential fit for w/ H.P.
Time [s]
BeII/ne ~ 0.13 %
Be/C ~ 17 sec
Be/C ~ 83 sec
Transient surface heating promotes Be2C formation leading to shorter mitigation times
Surface temperature during heat pulse ~ 1200°C[from R. Pungo et al., PSI17]
• Pulsing surface temperature to the 1200°C range results in faster chemical erosion suppression– Be2C disassociates at
~2200°C at 1 atm– Beryllium boiling point =
2471°C at 1 atm
• D retention during transient surface heating also increases by ~50% both with and without Be plasma seeding
U C S DU niversity o f C alifo rn ia S a n D iego
R. Doerner, IAEA CRP on H in Materials, Vienna, Sept. 26, 2006
0
500
1,000
1,500
2,000
0 5,000 10,000 15,000
Rea
ctio
n La
yer T
hick
ness
(nm
)
Time (seconds)
850 C
D 58 x 10-14
cm2sec
-1
750 C
D 0.43 x 10-14
cm2sec
-1950 C
Tungsten beryllide (BexW) formation may plague hot W plasma facing components
• Be2W and Be12W appear preferred (Be22W not seen)
• Beryllides only form in high temperature W surfaces (> 600°C)
• Be diffusion rate into W becomes significant above ~800°C
• At high temperature, Be availability (high vapor pressure of Be) on the surface can limit growth rate
Measurements from SNLL
U C S DU niversity o f C alifo rn ia S a n D iego
R. Doerner, IAEA CRP on H in Materials, Vienna, Sept. 26, 2006
Plasma conditions will play a dominant role in determining the impact of beryllides
• At high surface temperature, Be sublimation may prevent significant beryllide formation
• Sputtering at higher incident ion energies also tends to prevent significant beryllide formation
• Higher incident plasma flux tends to push energy, and temperature, necessary to avoid beryllides to larger values
T surf (K )
1100 1200 1300 1400 1500 1600
f Be+
10-3
10-2
10-1
10 eV
25 eV
50 eV
100 eV
D +=2 x1018 cm -2s -1
PISCES
D +=2 x1019 cm -2s -1
ITER
Uncoated Wsurface
Be coveredW surface
M. J. Baldwin et al., PSI17
fBeplasma(1–Rf) = YD-Beplasma(1–Rd) + fBeYBe-Beplasma(1–Rd) +evap(1–Re) + Dbulk
U C S DU niversity o f C alifo rn ia S a n D iego
R. Doerner, IAEA CRP on H in Materials, Vienna, Sept. 26, 2006
How might mixed materials impact ITER?
• Due to elevated temperature of C dump plates, carbides will likely form and limit C erosion
• If a full C divertor were employed, carbide formation on regions of the baffles, where the temperature is lower, would take longer, resulting in more C erosion and thereby more hard-to-remove tritium
• Be deposition on W baffles will likely not result in significant beryllide formation (TW ~ 400°C)
• If a full W divertor were used, beryllide formation near the strike points would be a concern (perhaps an issue for the JET ITER-like wall experiments)
• Beryllide formation in ITER only appears to be a concern on the W cassette liner ‘louvers’ (that are designed to be hot surfaces)