Report of the Special Expert Working Group on Chemical Erosion

1
Report of the Special Expert Working Groupon
Chemical Erosion

• S. Brezinsek

Institut für Plasmaphysik, Forschungszentrum
Jülich GmbH, EURATOM Association, Trilateral
Euregio Cluster, D-52425 Jülich, Germany
SEWG members CEA E. Tsitrone, E.
Delchambre FZJ A. Pospieszczyk, A. Kirschner,
A.Kreter, D. Borodin, V. Philipps IPP W.
Jacob, J. Roth, Ch. Hopf, T. Schwarz-Selinger R.
Pugno, A. Kallenbach, W. Bohmeyer, M. Baudach
UKAEA M.F. Stamp CIEMAT F. Tabares, D.
Tafalla, J.A. Ferreira FOM G. vanRooij,
J. Westerhout and TFE members
TF-E
2
Motivation - ITER
ITER
Research goal minimisation of risks and
optimisation of ITER availability!

• Lifetime issues
• Erosion, transport, and deposition
• of divertor material (CFC)
• Erosion, transport, and deposition
• of first wall material (Be)
• gt mixed material systems (Be, C, W)
• Control of transient heat loads
• (ELMs and disruptions)
• Safety issues
• Retention of tritium via co-deposition
• Methods to release the trapped tritium
• Dust formation (Be and T)

Beryllium
Tungsten

• Issues reflect the principle structure of
• the special expert working groups!
• All topics are related to each other!

Graphite
3
ITER Divertor Plasma Parameter Range
Result from AUG with W main chamber and C
divertor (Kallenbach et al. 2006) Outer
Divertor is the remaining source of carbon gt
Chemical erosion main process
B2-Eirene simulation for the ITER outer
divertor (A.S. Kukushkin)

• Standard operation scenario
• Semi-detached and recombining divertor plasma
  (Telt1.5eV)
• High ion and atom flux (Ggt31024 m-3)
• High surface temperature (up to1600 K)

4
Chemical Erosion Yield - Present Description

Hydrocarbon influx
G
chem
C
Fuel outflux
Erosion yield

• But in ITER
• Detached plasma operation
• Comparable atom and ion flux
• Low energy of incident particles
• Hot target surface

• Erosion yield Y as function of
• Surface temperature
• Incident ion energy
• Incident particle flux

5
Motivation - Tasks
We have to understand

• Which species are eroded from the PFC?
• Methane-, ethane-, propane family .
• Where are the species eroded from? How much is
  eroded?
• Spatial distribution of the erosion yield along
  the target .
• What does the plasma do with the eroded
  hydrocarbons?
• Hydrocarbon catabolism and transport.
• Where are the break-up products deposited?
• Inner divertor .
• Amount of deposited particles? Which hydrocarbon
  films are produced?
• Hard layers, soft layers.
• Which species is re-eroded?
• Higher hydrocarbons preferred .

Erosion
Migration
Deposition
Re-erosion
We should (try to) influence (control) these
processes to ensure a safe operation.
6
ITER predictions / SEWG
Predictions for ITER
Tokamak experiments JET, AUG, TEXTOR
Laboratory experiments MAJESTIX
Linear experiments pilot-MAGNUM, PSI-II
ERO modelling
Plasma background B2-Eirene or EDGE2D
Data base Codes HYDKIN ADAS, TRIM
Code and data base validation benchmark
experiments TEXTOR, AUG, TJ-II
7
Outline

• Recent experimental results from
• TEXTOR
• TJ-II
• ASDEX Upgrade
• JET
• Pilot-MAGNUM
• PSI-II
• Laboratory Experiments (IPP)
• ERO ITER predictions

8

• TEXTOR
• Photon efficiency calibration benchmark for
  HYDKIN and ERO
• Hydrocarbon catabolism
• 13CH4 experiments
• Long term deposition and erosion pattern
• .

9
Spectroscopic approach

• Photon flux proportional to particle flux (in
  ionising plasmas)

• Effective inverse photon efficiencies
  (D/XB-values) include dissociation chain
• D/XB-values depend on geometry, surface
  conditions etc.

• Chemical erosion yield is given by a corrected
  methane erosion yield

10
TEXTOR Photon Efficiencies for Methane
98032
98031
TEXTOR
TEXTOR

• D/XB values without contamination from deposited
  and re-eroded hydrocarbons
• Simple and highly reproducible plasmas

B
plasma
B
high
gas inlet
field
side
gas inlet
CD A-X
CD A-X
side view
top view
430.7/-1.0nm
430.7/-1.0nm
ERO modelling
Comparison with HYDKIN and ERO
calculations
CD A-X light
CD A-X band (CD from CD4)
D/XBExp. 36 D/XBHYDKIN46 D/XBERO32
11
TEXTOR Effective Photon Efficiencies for CxHy ,
CxDy
CH A-X band (Gerö band)
C2 d-a band (Swan band)
Brezinsek et al. PSI 2006
CxHy in D plasmas Te35 eV, ne2.21018m-3 at
the emission location CxDy in H plasmas Te45
eV, ne1.81018m-3 at the emission location
12
HYDKIN Photon Efficiencies for Methane
Comparison of measured effective photon
efficiencies with HYDKIN (Reaction kinetic
analysis - Reiter 2006) for CH from CH4 (const.
plasma, no deposition, no transport)
Brezinsek et al. PSI 2006

• Janev-Reiter database is reliable in the range
  between 5 eV and 100 eV
• Influence of other surface parameter and
  geometry loss of photons
• Data base is used as input for the ERO code!

13
TEXTOR CD Observed in Hot Edge Plasmas
gt CD identified in TEXTOR plasmas gt break-up
via molecular ions important

• TEXTOR benchmark experiment with CD4
• 20 of theoretically expected
• CI 909.5 nm per C observed
• 50 of theoretically expected
• CII at 426.7 nm per C observed
• 20 of theoretically expected
• Dg per D observed

14
HYDKIN CH vs. CH in the CxHy Catabolism
CH4
CH3
CH2
CH
C
C2
Part of the dissociation chain (CH4)
CH4
CH3
CH2
CH
C
Reaction kinetic analysis with HYDKIN (Reiter
2006) for CH4
Hot limiter plasma (TEXTOR, TJII)
Cold divertor plasma (AUG, JET)
CH
CH
CH
CH
Te45 eV
Te5 eV
15

• TJ-II
• Hydrocarbon injection experiments
• a-CH film production and destruction
• due to hydrocarbon puffing
• Nitrogen scavenger experiments
• .

16
TJ-II Hydrocarbon Injection Experiments

• Stellarator current free device Strong ripple
  ECRH heated plasmas
• Injection of CH4, C2H4 and H2 through a mobile
  instrumented graphite limiter
• Gas pulses of 12-15 ms (4 8x1018 particles)
• Much lower fuelling efficiency for C2H4 than for
  H2. Ha photon yield 15 per H atom that of H2
  (Garcia-Cortes et al. JNM 2005)
• Photon flux ratio CH/Ha 3 times higher for C2H4
  than for CH4 (Te3eV ne2x1018 m-3) ? deposition
  of carbon films with high H content

Ha
CH
Tabares et al. PSI2006
C2H4
17

• AUG
• Photon efficiencies, hydrocarbon particle fluxes
  
• and erosion yield in detached and attached
  plasmas
• Long term erosion deposition pattern
• Carbon source determination
• 13CH4 injection experiments
• .

18
AUG Hydrocarbon Injection in Detached Plasmas

• L-mode discharges with outer divertor
  detachment
• Injection of CH4 and C2H4 in the SOL,
  Separatrix and PFR

in the volume Te1.2 eV, ne31020m-3
at the target (separatrix) Te2.3 eV,
ne41018m-3
19
AUG Photon Efficiencies in Detached Plasmas
Intrinsic CD A-X spectrum --- attached plasma ---
Intrinsic CD A-X spectrum with strong BD
contamination --- detached plasma ---
CH A-X spectrum from injection and CD A-X
intrinsic background --- detached plasma ---
CH A-X spectrum from injection and CD A-X
intrinsic background --- attached plasma ---
Brezinsek et al. PFCM 2006
20
AUG Photon Efficiencies in Detached Plasmas
Intrinsic C2 d-a spectrum --- attached plasma ---
Intrinsic C2 d-a spectrum --- detached plasma ---
Below the detection limit!
C2 d-a spectrum from injection --- detached
plasma ---
C2 d-a spectrum from injection --- attached
plasma ---
Brezinsek et al. PFCM 2006
21
AUG Erosion Yields in Detached Plasmas
Measured D/XB values are slightly higher than in
attached plasmas and higher than predictions with
HYDKIN!
D/XB for CH from CH4 18 /-7 D/XB for CH from
C2H4 47 /-19
Decrease of CH and C2 light much stronger than
increase of D/XB gt Significant decrease of the
hydrocarbon flux in detachment
Erosion yield at the separatrix for detached
conditions
Ychem3.210-2 (ion flux only) Ychem2.910-3
(atom and ion flux)
Roth-formula Ychem2.510-3 (atom and ion flux)
Reduced influx is largely compensated by lower
plasma ion outflux gt atom outflux important
22

• JET
• Hydrocarbon injection experiments
• a-CH film production and destruction
• as function of the strike-point
• Nitrogen scavenger experiments
• 13CH4 tracer experiment
• .

23
JET CxHy Contribution to the Erosion Yield
Circumferential injection into the attached outer
divertor i) Discharge with C2H4 injection ii)
Discharge with H2 injection Nearly identical
local plasma conditions
Assumption symmetric and homogenous injection
24
JET CxHy Contribution to the Erosion Yield

• Most reliable value for Ychem is achieved
• when the strike point is at GIM10
• Toroidal inhomogeneity of gas injection
• module included (gt information from
• 13CH4 tracer experiments)
• Bypass in the outer divertor considered!
• Photon efficiencies and erosion yield
• lowered in comparison to first analysis
• (Brezinsek et al. EPS 2005)

• D/XB for C2 Swan band from C2H4 about 75
• Ychem associated to higher hydrocarbons about
  0.6

Brezinsek et al. PSI 2006
25
JET Preferred Release of Hydrogen-rich
Hydrocarbons

• Ratio of C2/CH line emission indicate
  hydrogen-rich layer near to louvre
• Strike-point utilised to clean up deposited
  areas
• Thermal decomposition of deposited layers most
  probable

Brezinsek et al JNM2005

• Spectroscopy detects strong C2 light emission
• Deposition monitor strong material deposition

Soft layer detected and (re)moved/cleaned.
26
JET History Effect

• Material deposition on the QMB depends on the
  strike-point position of
• previous discharges!
• Identical plasma discharges lead to different
  net deposition on QMB
• Strike-point configuration as well as ELM type
  influence deposition

27

• Pilot-MAGNUM
• Erosion yield at ITER-relevant fluxes
• .

28
Pilot-MAGNUM Plasma Parameter Range
Coils
Ion fluxes and plasma parameters at the target
are comparable to predicted ITER divertor
conditions!
0.55 m
Target
Arc
gt1023 H/m2s
gt1024 H/m2s
gt1025 H/m2s
Thomson Scattering
Spectroscopy

• Parameter variation
• Source current flow
• Vessel pressure
• Magnetic field
• Target potential

v. Rooij et al. ITPA Toronto 2006
29
Pilot-Magnum First Results on Erosion Yields
Ionising hydrogen plasma (Te3.5 eV) in front of
the fine grain graphite target.

• Preliminary work based
• on CH A-X band analysis
• gas injection for
• calibration in preparation
• uncertainties in the
• surface temperature
• New experiments planed!

Experimental data in-line with the Roth-formula!
30

• PSI-II
• Re-erosion of hydrocarbons at elevated wall
  temperatures
• Hydrocarbon spectroscopy
• .

31
PSI-II CxHy Erosion from Deposited a-CH Layers

• Injection of methane into a PSI-II hydrogen
  plasmas
• a-CH layers build up on the heatable wall
• Erosion by neutral hydrogen
• Detection of the eroded species with the aid of
  QMS and
• collector probes in the pump duct

TMP
QMS analyses the exhaust 1 m upstream forepump
target chamber
duct
discharge region
20 cm
CH4 100 sccm H2
heated
Collector probes for in situ analysis
200C
150C
100 sccm H2
0.8 Pa
0.1 Pa
Bohmeyer et al. PFCM 2006
Large collector probe
Langmuir probe
TM pumps
floating W disk
32
PSI-II CxHy Erosion from a-CH Layers at Walls
Increasing wall temperature
Wall temperature fixed at 180 C
QMS in the pump duct
Dominant erosion product from the hot wall
(180C) is acetylen (m26)!
33
PSI-II Deposition/Erosion in the PSI-II Pump Duct

• Unheated duct deposition ( 0.010.05 nm/min)
  along the duct (2 sccm CH4)

• Heated collector probe erosion (2 sccm CH4 )

Erosion by atomic hydrogen
Reduced erosion during CH4 injection
Bohmeyer et al.

• Surface temperature is the key parameter for the
  erosion/deposition switch

34

• Laboratory experiments (IPP Garching)
• Nitrogen/hydrogen mixtures
• Re-deposition of soft a-CH films after thermal
  decomposition
• .

35
Redeposition of Soft a-CH Films after Thermal
Decomposition
TESS Setup
Jacob et al. 2005
36
TESS Thermal effusion spectra
Jacob et al. 2005
Tramp 15 K/min

• - Large contributions from
• long chain hydrocarbons
• High redeposition (5ß)

37
TESS Redepositon
Shift of oven causes desorption of material which
was redeposited on the walls of the glass
tube. Redeposition fraction from integration ? 41
38

• ITER
• Predictive modelling for ITER with ERO (here
  only C)

39
ITER Chemical Erosion Yield at the Outer Target

• Chemical erosion yield (D? graphite) along outer
  target

• Erosion yield YRoth - min(YRoth) 0.04
  around strike point (high flux)
• - max(YRoth) 0.6 _at_ Tsurf 670C

40
ITER predictions / ERO
Present status (Kirschner et al. 2006)

• no lifetime problem
• (1mm in 500 ITER discharges)
• T co-deposition critical for D phase
• (30 ITER discharges)
• gt not an issue for H phase

Still large uncertainties! Improvements possible!

• Influence of Be
• C erosion reduced by a factor of 2.5
• T co-deposition problem remains
• SEWG ITER-material mix / T retention

Implementation of new results will help to
increase the level of confidence!
41
Summary
Predictions for ITER
Tokamak experiments JET, AUG, TEXTOR
Laboratory experiments MAJESTIX
Linear experiments pilot-MAGNUM, PSI-II
ERO modelling
Plasma background B2-Eirene or EDGE2D
Data base Codes HYDKIN ADAS, TRIM
Code and data base validation benchmark
experiments TEXTOR, AUG, TJ-II
SEWG meeting in MARCH 2007 at JET!
42
MERO Matlab GUI for ERO data processing and
visualisation
43
PSI-II C2H4 injection into He-Plasmas
CH A-X band
CH A-X band
plasma column
ne1017m-3
ne31017m-3
HYDKIN cannot reproduce the short penetration
depth for CH in PSI-II. Uncertainties at low
electron temperatures!
CH A-X band
CH A-X band
ne61017m-3
ne21018m-3
gas injection
gas injection
Baudach et al. PFCM 2006
44
Plasma Parameters in the Inner and Outer JET
Divertor
45
Predictive ERO modelling for ITER conclusions

• Estimations of upper limits of tritium
  retention
• - 0.1 Be 6.4 mg T/s (350 g after 150 ITER
  discharges)
• - 1.0 Be 15.9 mg T/s (60 ITER discharges)
• - 1.0 C 32 mg T/s (30 ITER discharges)
• ? Target lifetime seems to be less critical
• - worst case (0.1 Be) 6900 ITER discharges
• Open questions
• - influence of disruptions and ELMs
• - carbide formation (Be2C)
• - T content in dependence on surface temperature
• - T retention in gaps

46
TJ-II Hydrocarbon Injection Experiments
Summary Gas Injection
47
AUG Plasma Parameters at the Target