SEWG MEETING, Cadarache, June 16th 2009 1/18 Experimental studies on laser induced removal and collection for absorbing particles A.Vatry 1,2, A. Marchand.

SEWG MEETING, Cadarache, June 16th 2009 1/18

Experimental studies on laser induced removal and collection for absorbing particles

A.Vatry1,2, A. Marchand1, Ph. Delaporte1, C. Grisolia2, M. Sentis1,

C. Hernandez2, H. Roche2

1 Lasers, Plasmas and Photonic Processes Laboratory, Marseille, Francewww.lp3.univ-mrs.fr

2 Association Euratom/CEA, DRFC/SIPP, Saint Paul lez Durance, France

http://www.lp3.univ-mrs.fr/

http://www.cea.fr/


MOTIVATIONS

The dust have high adhesion force the laser is a good solution for the removal

This is the first step for the cleaning

Decontamination of fuel tubes (AREVA, ONET, CEA) Microelectronics

Particles : Moxsubstrate : zircalloy

Optics

Particles : polymersubstrate : silicon

Particles : organic, metallicsubstrate : silica

Removal technique based on laser cleaning are developed in various field

Advantages : handle remote, free contact

http://www.cea.fr/


Experimental set-up

aperture

Attenuating plate lens

laser

Examples of optical microscope images of C particles on Si substrate:

N0 particles

Substrate with particles

dVacuum chamber

N particles

PRE set-up

Before the irradiation After 5 shots at 200mJ/cm²

aperture

Attenuating plate lens

Laser

Collector substrate

Collection set-up

Collector substrates have been observed with

an optical and scanning electron microscopes

4 ns – 1064 nm

0

1N

NPRE

http://www.cea.fr/


lens

XeCl laser

target

substrate

KrF

laser

▪ Dusts are produced by laser

ablation of graphite or W target and

are collected on substrate

The laser-produced dusts have

similar morphologies and nature

than Tokamak ones.

Tokamak and Argon discharge

Laser at LP3

200nm

Experimental set-upTextor

ASDEX Upgrade (PSI poster Balden)

DC Argon discharge

C produced under vacuum (10-2mbar)

W produced under air

C produced under air

Dusts production

sample

http://www.cea.fr/


Carbon particles produced by laser

200nm1µm

Produced under air:

Produced under low pressure:

C on Si produced under 10 mbar d’He

▪ Layer and particles

▪ sizes ~ < 1 µm

▪ Very porous

▪ Amorphous structure

900

800

700

600

500

400

300

1600140012001000800

Bande D

Bande GI (u.a)

Raman shift (cm-1)

Raman spectrum achieved in PIIM laboratory(Cédric Pardanau)

1µm

▪ Separated particles

▪ Various shapes

▪ sizes ~ < 1 µm

▪ Very porous


PRE for carbon particles

4 ns – Si substrate

Wavelength influence:

▪ Laser is very efficient to remove carbon particles, for F> 500 mJ/cm² More than 80% of particles are removed

▪ Wavelength has a low influence on the PRE

▪ Pulse duration has low influence, except for a long pulse duration as 200 ns

Influence of laser parameters :

Pulse duration influence:

0

0,2

0,4

0,6

0,8

1

0 200 400 600 800 1000 1200Fluence (mJ/cm²)

PR

E f

or 5

sho

ts

450 fs

50 ps

7 ns

At 200 ns - 1064 nm PRE = 0 for fluences up to 1.16 J/cm²

1025 nm-1064 nm

http://www.cea.fr/


PRE for carbon particles

Si and SiO2 have very different thermal and optical properties

However the curves are quite similar

PRE and then removal mechanism do not depend on the substrate properties

0

0,2

0,4

0,6

0,8

1

0 100 200 300 400 500 600 700 800

Fluence (mJ/cm²)

PR

E f

or

5 s

ho

ts

C on Si

C on SiO2

50 ns – 308 nm

Substrate influence:

0,0

0,2

0,4

0,6

0,8

1,0

0 200 400 600 800 1000 1200

Fluence (mJ/cm²)

PR

E fo

r 5

shot

s

C on SiO2

C on Si

4 ns – 1064 nm


Collection of carbon particles

Almost no intact particles are collected

The carbon particles are ablated during the irradiation

Collector substrate for d = 5 mm, under 10-3mbar, for 1 shot F = 380 mJ/cm²

The ejected carbon particles are in the form of very thin nanoparticles or/and atoms

KrF laser: 248 nm – 27 ns

Carbon particles produced by laser ablation

http://www.cea.fr/


Removal mechanism of carbon particles

▪ The removal mechanism is the particle ablation, due to the direct absorption of the laser energy by the particle▪ If the particle is too big, the removal is achieved in several steps


Comparison with dust from Tore Supra

Carbon particles produced by laser ablation Dust collected in Tore Supra

Laser is also efficient for the particle collected in Tore SupraThe difference of efficiency is explained by the difference of size

4 ns – Si substrate 248 nm– 27 ns – Si substrate


Carbon dust on CFC tile

248 nm - 27 ns - 300 mJ/cm²

Before irradiation 5 shots

Sample achievement:

Result:

A laser is focused on a CFC tile

Particles are deposited around the crater

We irradiate the tile with a laser at 300 mJ/cm²CFC tile


▪ Particles produced by PLD with tungsten target 2 kinds of particles

► Very thin aggregates (foam)

► Droplets of 1 to 5 µm with very smooth surface

Tungsten particles produced by laser

W droplets are also produced in plasma device or by arcing in Tokamak

Produced by plasma gun

Collected in ASDEX Upgrade

Produced by laser


1064 nm - 4 ns - ~700 mJ/cm²248 nm - 27 ns - ~800 mJ/cm²

▪ UV range is very efficient to remove the very thin aggregates

0 shot

1 shot

Removal efficiency for W thin aggregates

▪ Infrared range do not remove efficiently thin aggregates


Laser could be efficient to remove tungsten particles but for specific parameters

Wavelength has a great influence on PRE

UV beam is more efficient than infrared one

PRE of W droplets

The pulse duration has an influence too

But the damage threshold of the substrate is very low we cannot use picosecond and femtosecond pulse duration for our application

4 ns – Si substrate

Wavelength influence:

Influence of laser parameters :


At 200 ns - 1064 nm PRE = 0 for fluence up to 1,25 J/cm²

http://www.cea.fr/


Collection of tungsten particles

Collection:

Collector substrate for d = 5 mm, under 2.10-2 mbar, 1 shot, F = 700 mJ/cm²

Collector substrate for d = 3 mm, under air, 5 shots, F = 700 mJ/cm²

A lot of intact particles have been collected under air and primary vacuum

KrF laser

248 nm – 27 ns

http://www.cea.fr/


1000

800

600

400

200

0

806040200

2000

1500

1000

500

Temperature (K)

Time (ns)

FORTRAN 1D simulation :

Temperature below the fusion threshold

Studies on tungsten particles removal

30 shots5 shots

4 ns – 1064 nm – ~700 mJ/cm²

No particle damage

Collection results are in accordance to FORTRAN simulation Particles could not be ablated

1000

800

600

400

200

0

806040200

2000

1500

1000

500

Depth (µm)

5 shots1 shots No particle damage

27 ns – 248 nm – ~800 mJ/cm²

1000

800

600

400

200

0

806040200

1200

1000

800

600

400

Depth (µm) Temperature (K)

Time (ns)

1000

800

600

400

200

0

100806040200

1200

1000

800

600

400

Particles not removed Particles not removed



248 nm - 27 ns - 500 mJ/cm²

Before irradiation 5 shots

5 µm

Tungsten dust on CFC tile

Sample achievement:

Result:

lens

XeCl laser

Tungsten target

KrF

laser

CFC tile

A laser is focused on a W target

Particles are deposited on the CFC tile


Summary

The laser is efficient to remove dust

For tungsten dust specific parameters are required

The carbon dust are ablated collection of atom and/or very thin nanoparticles

The tungsten droplets are ejected intact collection of micrometer particles


Thank you for your attention


PRE for W droplet

Substrate influence:

0

0,1

0,2

0,3

0,4

0,5

0,6

0 500 1000 1500 2000

Fluence (mJ/cm²)

PR

E

W on Si

W on SiO2

50 ns – 308 nm

The substrate can has an influence

However, in both case, for absorbent and transparent substrate the laser can remove carbon particle

50 ps – 355 nm


1000

800

600

400

200

0

806040200

8000

6000

4000

20001000

800

600

400

200

0

806040200

2000

1500

1000

500

Temperature (K)

Time (ns)

FORTRAN 1D simulation:



Temperature (K)

Time (ns)

Depth (µm)

Temperature above the fusion and vaporization threshold

Vaporization (5828 K)

Fusion (3695 K)

30 shots5 shots

4 ns – 1064 nm – ~700 mJ/cm²

No particle damage 5 shots

1 µm

Particle damage


1000

800

600

400

200

0

806040200

2000

1500

1000

500

1000

800

600

400

200

0

100806040200

8000

6000

4000

2000

Depth (µm)

50 ps – 1064 nm – 540 mJ/cm²




248 nm – 27 ns – 800 mJ/cm²

1025 nm – 450 fs – 300 mJ/cm²355nm – 50 ps − 290 mJ/cm²

5 shots4 shots

0 shot 1 shot

1 shot

Under the removed particle the substrate is not affected by the laser

SEM observations:

http://www.cea.fr/


Field modification around the W particles :

7 ns – 1064 nm – ~ 800 mJ/cm² 450 fs – 1025 ns – ~300 mJ/cm²355nm – 50 ps − 290 mJ/cm²

15 shots5 shots 15 shots

Simulation made by Nicolas Bonod (Fresnel Institute)

▪ The Si texturization allows the materialization

of the field around the particle

▪ The interface is in the particles shadow

▪ No interface mechanism could be at the origin

of the removal


5µm

Possible removal mechanism :

▪ Particle ablation

▪ Explosive evaporation of humidity

▪ Thermally induced expansion

▪ Local substrate ablation

▪ Electrostatic force

http://www.cea.fr/



First calculations about electrostatic force (266 nm- 4ns- 800 mJ/cm²):

Comparison between Van der Waals force and Electric force :

Force between positive charge of 5 µm particles and photo-electrons at 4ns :

F = ~ 2x10-7 N

The adhesion force between a metal particle (with

deformation) and the silicon subtrate is ~10-6 or 10-7 N

This is the same order of magnitude

Pulse duration influence :

For picosecond and femtosecond regime

multiphotonic effect

More photoelectrons

Higher removal efficiency


Si substrate

http://www.cea.fr/


Particle removal mechanism

Thermally induced expansion Local substrate ablation

Explosive evaporation

Destruction of substrateMechanical process

temporal parameters

Adsorbed humidity

Particle ablation

depends on materials

Electrostatic force

▪ Different physical

mechanisms could lead to

the particle removal

▪ Several mechanisms could

be involved simultaneously

in the removal

http://www.cea.fr/


0

0,2

0,4

0,6

0,8

1

0 200 400 600 800 1000 1200Fluence

PR

E

450 fs

50 ps

7 ns

Influence des paramètres laser (sur Si):

Durée d’impulsion :

À 200 ns - 1064 nm PRE = 0 pour F = 1,16 J/cm²

1025 nm-1064 nm



À 200 ns - 1064 nm PRE = 00

0,1

0,2

0,3

0,4

0,5

0,6

0 200 400 600 800 1000 1200

Fluence (mJ/cm²)

PR

E à

5 t

irs

450 fs-1025 nm

50 ps-1064 nm

4 ns-1064 nm

À 200 ns - 1064 nm

PRE = 0, jusqu’à F = 1,25 J/cm²


1000

800

600

400

200

0

806040200

8000

6000

4000

20001000

800

600

400

200

0

806040200

2000

1500

1000

500

Temperature (K)

Time (ns)

FORTRAN 1D simulation:



Temperature (K)

Time (ns)

Depth (µm)

30 shots5 shots

4 ns – 1064 nm – ~700 mJ/cm²

No particle damage 5 shots

1 µm

Particle damage


1000

800

600

400

200

0

806040200

2000

1500

1000

500

1000

800

600

400

200

0

100806040200

8000

6000

4000

2000

Depth (µm)

50 ps – 1064 nm – 540 mJ/cm²


1000

800

600

400

200

0

806040200

8000

6000

4000

2000

2

Fusion (3695 K)

Vaporisation (5828 K)

Temperature above the fusion and vaporization threshold

Possible removal mechanism :

▪ Particle ablation

▪ Explosive evaporation of humidity

▪ Thermally induced expansion

▪ Local substrate ablation

▪ Electrostatic force

http://www.cea.fr/


1000

800

600

400

200

0

806040200

2000

1500

1000

500

Temperature (K)

Time (ns)

FORTRAN 1D simulation :



30 shots5 shots

4 ns – 1064 nm – ~700 mJ/cm²

No particle damage


1000

800

600

400

200

0

806040200

2000

1500

1000

500

Depth (µm)

5 shots1 shots No particle damage

27 ns – 248 nm – ~800 mJ/cm²

1000

800

600

400

200

0

806040200

1200

1000

800

600

400

Depth (µm) Temperature (K)

Time (ns)

1000

800

600

400

200

0

100806040200

1200

1000

800

600

400


http://www.cea.fr/


C particles at 0.1 Pa, F = 1.6J/cm²

0 500 1000 1500 2000 2500 30000

1

2

3

4

5

6

7

8

9

10

11

12

Si SiO2

d (

mm

)

t (ns)

v = 3800 m.s-1

v = 4240 m.s-1

C particles on Si substrate at 0.1 Pa

(velocity slope of the linear regression curve for 1 µs )

v = 3150 m.s-1

v = 3920 m.s-1

v = 4390 m.s-1

0 200 400 600 800 1000 1200 1400 16000

1

2

3

4

5

6

700 1 1.6

dis

tan

ce (

mm

)

time (ns)

mJ/cm²

J/cm²J/cm²

time (ns)

Dynamic of ejection for carbon particle

Substrate influence: Fluence influence:

▪ These graphs are performed with pictures of the light emitted by the ejected particle or ablated species take with a CCD intensified


0

0,1

0,2

0,3

0,4

0,5

0,6

0 200 400 600 800 1000 1200

Fluence (mJ/cm²)

PR

E à

5 t

irs

450 fs-1025 nm

50 ps-1064 nm

7 ns-1064 nm

Efficacité d’enlèvement – particules de tungstène



À 200 ns - 1064 nm

PRE = 0, jusqu’à F = 1,25 J/cm²

0 tir 5 tirs

1064 nm – 200 ns - 625 mJ/cm²

1064 nm - 4 ns - ~700 mJ/cm²

1025 nm - 450 fs - ~300 mJ/cm²

0 tir 5 tirs


Outlook

LASK V1

P=10-6PaT= 200°C max

Environment :

Dust collection : adhesion

LASK V2

P=AtmT= 50°C

Environment :

Dust collection : aspiration

These studies help to the elaboration of a removal device based on laser cleaning for Tore Supra. This is a first step for the ITER removal device

http://www.cea.fr/

SEWG MEETING, Cadarache, June 16th 2009 1/18 Experimental studies on laser induced removal and collection for absorbing particles A.Vatry 1,2, A. Marchand.

Documents

particles substrate

laser slide

ns carbon particles

carbon particles si

particles sizes

silicon particles

kinds of particles

absorbing particles