C.H. Skinner, H.W. Kugel, Princeton Plasma Physics Laboratory , J. T. Hogan, Oak Ridge National Laboratory, W.R Wampler, Sandia National Laboratories, and the NSTX research team.

Deposition measurements in NSTX

Abstract

Two quartz microbalances have been used to record deposition on the National Spherical Torus Experiment. The experimental configuration mimics a typical diagnostic window or mirror. An RS232 link was used to acquire the quartz crystal frequency and the deposited thickness was recorded continuously with 0.01 nm resolution. Nuclear Reaction Analysis of the deposit was consistent with the measurement of the total deposited mass from the change in crystal frequency. We will present measurements of the variation of deposition with plasma conditions. The transport of carbon impurities in NSTX has been modelled with the BBQ code. Preliminary calculations indicated a negligible fraction of carbon generated at the divertor plates in quiescent discharges directly reaches the outer wall, and that transient events are responsible for the deposition.

Supported by DOE Contract No. DE-AC02-76-CH03073.

Deposition on TFTR graphite tile

S Willms and W Reisiwig LANL

50 µm

Motivation:

• Condition of plasma facing surfaces is key factor in plasma performance, but generally remains hidden and undiagnosed.

• Deposition: – affects recycling, – covers up boronized layer– coats diagnostic windows & mirrors– is major cause of tritium retention

• For 2005 NSTX will reconfigure boronization probe and has embarked on lithium wall conditioning by pellets / evaporation / and possibly a flowing Li wall module.

• Real-time direct measurements of deposition are crucial to obtain understanding and control of deposition.

Plasma facing components in contact with the plasma are protected by graphite and CFC tiles.

0.1

1

10

0 50 100 150 200 250 300 350

carbondeuteriumboron

100

1000

areal density 1e18/cm

2

toroidal angle (degrees)

nm

at density of 1.6 g/cm

3

NSTX Deposition on Wall Coupons

• Factor of 5 Variation in Deposition Toroidally

• Factor of 6 to 9 Variation in Deposition

Poloidally

Deposition from plasma operations September

2000

to August 2001 including 9 boronizations

gas injectors0.01

0.1

1

10

-60 -40 -20 0 20 40 60

carbondeuteriumboron

10

100

1000

1e18 atoms / cm

2

poloidal angle (degrees, from outboard midplane)

nm

at density of 1.6 g/cm

3

film mass frequency change

crystal mass bare crystal frequency=

Crystal ‘wobbles’ in thickness shear mode

Crystal frequency spectrum

Inficon

Inficon

More complex relation available that takes account of acoustic properties of film and is valid up for to 40% frequency shifts.

Calculated sensitivity: -81 Hz / µg /cm2

or -13 Hz /n.m. (for film density of 1.6 g/cm3)

Caveat:crystal frequency also sensitive to temperature,.

5.9 MHz fundamental resonance constantly identified by zero phase difference between applied signal voltage and current passing through crystal.

Quartz Crystal Microbalance

Deposition changes crystal oscillating frequency.• Widely used for process control during vacuum deposition.• Used in TdeV, ASDEX, JET, TEXTOR and NSTX tokamaks. • Frequency measured by pulse accumulator controlled by 20 MHz reference oscillator.• Commercially available system (Infincon XTM/2) is relatively fast (1/4 s), precise, able to measure heavily loaded crystals and has immunity from mode hopping

Two deposition monitors are installed on NSTX

detectorinside

Crystal @ R= 231 cm, ~ 80 cm from last closed flux surface, 33 cm below midplane

mimics typical diagnostic windows and mirrors,

- samples neutrals deposited on NSTX wall

oscillator

NSTX Bay K

Gold coated quartz crystal

View from plasma side:

Si witness coupon

Gold coated quartz crystal

Thermo-couple

Shutter (folded back)

Rear facing crystal

4” port

•Quartz crystal oscillates at ~ 5.9 MHz, the exact frequency depends on mass and on temperature.

•Deposition inferred from change in frequency (measured to ~0.1 Å, ~0.05 Hz)

•Location relatively far from plasma enables continuous recording of deposition before and after discharge.

•1/4 s response time

•Thermocouples on both detectors record temperature

•Bakeable to 400°C.

•Built in shutter on plasma facing detector.

•Data acquisition via analog port and RS232 link

• Commercially available, Inficon XTM/2

Deposition over period January 10th - February 14th 2003Continuous rise punctuated by sharp material loss.

0

5

10

15

20

109500 109700 109900 110100

10

30

50

70

90

110

130

deposition (µg/cm

2 )

shot number

deposition (nm)

Deposition Results 2003:

Excellent agreement between mass measured by nuclear reaction analysis and by quartz microbalance

Deposition on back facing crystal 1.30 µg/cm2,

- about 10% of front facing crystal

Data reflects sticking probability of hydrocarbon

radicals.

1.28

1.32

1.36

25

26

14.4 14.8 15.2 15.6 16

µg/cm

2

degrees C

Time of day (decimal hours)

4

56

7

Deposition often occured at first discharges in day.

Typically the first shots in the day showed deposition.The temperature (blue curve) shows that the slow continuous rise in deposition is not a temperature effect

temperature

deposition

• 2004 data acquired continuously at higher resolution via RS232:

• Slow deposition continues between discharges[1].

• Important implications for migration of codeposited tritium

• NSTX diagnostic shutters now closed imeadiately at end of discharge to minimise exposure.

Deposition over 4 shots 112014-017

1Å

Average net deposition in 2003 0.027 µg / dischargein 2004 0.005 µg / discharge

[1] A. von Keudell, C. Hopf, T. Schwarz-Selinger, W. Jacob, Nucl. Fus, 39 (1999) 1451. “It is also conceivable that, due to the relatively high surface temperature in JET (>500K), polymer-like films produced during plasma interaction evaporate after each discharge and are deposited on the cold louvers in pulse pauses as well.”

20040409

Deposition shows slow but persistent rise during plasma ops.

•Calibrate temperature response of crystal frequency with heated water.•Convert temperature and crystal frequency to equivalent Å. •Rear crystal frequency tracks temperature•Front crystal frequency shows excess due to continuous slow deposition.

Modelling suggests ‘bursty’ transport responsible for deposition

BBQ details:Collision model Similar to LIM, WBC, ERO codes, detailed magnetic geometry (EFIT)

Background parameters assumed (ne, ln, Te, lT)[ local D+ flux amplification, sheath electrostatic (es) field, ESOL ] MZ dVZ/dt = -Ffriction -Fes + random // and ^ diffusionFZ = -MZ (VSOL-VZ) / ts ; Fes = Ze ESOL F// = Random // diffusion, D//=(8EZ/3pMi) t //

F^ = Random ^ anomalous diffusion ( D ^ )Particle energy (WZ)dWZ/dt = (kTi -WZ)/tT +Rfriction +Res

Molecular processes: Erhardt-Langer,Janev-Reiter hydrocarbon break-up ratesAtomic processes (ionization, recombination, D0 charge exchange) for carbonBirth gyro-collisions with surface

-1.60

-1.50

-1.40

-1.30

0.30 0.40 0.50 0.60

-1.633.441.810.185

R (m)

1.63

0

0

0.333

0.666

1Z (m) Localization of neutralC density near strike points

Configuration from EFITNSTX pulse 1108251 @ 235ms

CI density(rel.)

θ

Poloidal localization at strike points

• BBQ calculations of quiescent cross-field transport of impurities generated at the divertor strike points find little mid-plane deposition, using conventional models.

• ‘Bursty’ low-field side, far-SOL transport and/or ELMs should give higher deposition rates

John Hogan ORNL

Boronization measured directly

Note sensitive tracking of few monolayers only of boron deposition processes and apparent subsequent removal of deuterium on plasma facing crystal (well correlated with logbook entries).

Total thickness ~5Å very low due to poor penetration of glow discharge to detector.

Future: Plan to use three detectors at upper, lower divertor and midplane to monitor B and Li deposition directly.

Note: • Crystal frequency also changes due to temperature change.• Plot shows thickness after averaged temperature compensation• Crystal measures deposited mass.• Thickness derived from assumed density of 1.6 g/cm3

50

55

60

65

-65

-60

-55

-50

8 10 12 14 16

PERSEC_TAB.20040424

Time of Day

Deposition - rear facing crystal

Deposition - plasma facing crystal

Start He GDC

Start TMB

"TMB flow problems"

Finish TMB

Finish He GDC

"He flow problems"

Boron deposition

Deuteriumremoval

Deposition on plasma facing and rear facing crystals during Saturday boronization 24 April 2004.- comments from operators logbook.

(zero point arbitrary)

• NSTX shows net deposition on wall, heaviest at midplane and near gas injectors

• 29 µg/cm2 deposition accumulated over 4 weeks and 497 discharges with 16 µg/cm2 material

loss over 7 discharges 77 cm from plasma in diagnostic mirror-like geometry.

• Mass of deposited layer independently confirmed by ion beam analysis.

• Deposited layers are mostly carbon with some boron, oxygen and deuterium.

• Deposition recorded on surface facing away from plasma from low sticking

probability radicals at 10% rate of deposition on plasma facing surface.

• Deposition typically apparent on first discharges of day.

• Deposition slowly accumulates between discharges - H-isotope migration ?

• Deposition measured during boronization.

• Monte Carlo modeling suggests that transient processes are likely to dominate the deposition.

Summary

50

55

60

65

95

100

105

110

0 5 10 15 20Time of Day

Crystal2 frequency

Crystal2 temp.

20040708<---- Plasma Ops. ----->

Rear facing crystal

-20

-15

-10

95

100

105

110

0 5 10 15 20

Å

equivalent Å

Time of Day

Crystal1 frequency

Crystal1 temp.

20040708<---- Plasma Ops. ----->

Plasma facing crystal

25

30

35

40

25

30

35

40

0 5 10 15 20

(Å)

(Å)

Time of Day

Plasma facing

Rear20040708

Deposition after temperature compensation

<---- Plasma Ops. ----->

46th Annual Meeting of the Division of Plasma Physics, 15-19th November 2004, Savannah, Ga

deuterium (IBA) 0.20 µg/cm2

carbon (IBA) 11.4 µg/cm2

oxygen (from XPS)

1.96 µg/cm2

D+C+O mass 13.5 µg/cm2

Microbalance 13.3 µg/cm2

carbon 73%

boron 9%

deuterium 18%

Deposit composition fromion beam analysis 2004

QuickTime™ and aNone decompressor

are needed to see this picture.

Side view

Si coupon

Plasma facing crystal

Backwardfacingcrystal

V.V.

wall

Top view

Plasma

LCFS

0

5

10

15

20

25

0 50 100 150 200 250 300 350

C neutral

C+

C2+

C3+

C4+

C5+

C6+

C totalC density (rel.)

poloidal angle

outerstrike point

innerstrike point