Transcript ppt

A comparison of optical trains based on a GIMM
or a Dielectric Mirror final optic
HAPL San Diego, 9th Aug. 2006
Malcolm W. McGeoch
PLEX LLC
280 Albany St.
Cambridge MA 02139
617-621-6300
Baseline HAPL final optical parameters:
1. 2.5MJ at 5Hz, 40 illumination beams each 62.5kJ
2. 2 Jcm-2 in optical distribution ducts.
3. Duct aspect ratio 6:1, each beam 3x18 beamlets
(area of one beam = 3x18x(0.24)2 = 3.1m2)
4. Focal length 39m (GIMM) or 42m (all-Dielectric case)
5. Vertical “slits” in blanket, total 0.6% of 4p
(slit size 1.35m high x 0.22m wide for GIMM case)
24cm x 24cm beamlet
from de-multiplex array
The final optics must function to attenuate the neutron flux:
target
3x1024 /beamline /FPY
n
laser building
< 1x1019 /beamline /FPY
= < 1mrem/hr/day
at one day after shutdown
(M. Sawan, this meeting)
Attenuation factor of 106 !
We can encase the optical path with 1.5m - 3m of concrete
and attenuate using the angles that are needed in any case
for distribution around the sphere
target
3x1024 neutrons/beamline /FPY
n
< 1x1019 /beamline /FPY
Concrete + ferritic steel + H2O coolant
40 port arrangement: 3 tiers, 8 longitudes each hemisphere
View from above “North Pole”
Elevation
Uniformity for 9-element “basket”: N = 3,4,5 with g = 0.8, 0.9, 1.0
Mean uniformity of basket (% rms)
2.0
icosahedron
30 beam 3-6-6 + offset
40 beam 4-8-8 + offset
50 beam 5-10-10 +offset
60 beam 6-12-12 + offset
truncated icosahedron
1.5
72 beam 6-6-12-12 + offset
all 3-tier cases have
1.0
24 / 52 / 79
o
polar angles
0.5
0.0
20
30
40
50
60
Number of Beams
70
80
Elevation sketch of beams for all-dielectric final optics
turning mirror M3
plane mirror M1
(3J cm-2)
focusing mirror M2
(hollow)
main containment wall
22.5m
20m
blanket
10.75m
32.5m
66 deg
38 deg
11.85m
11 deg
"Ground Level"
entry mirrors (turn beams from
horizontal to vertical propagation)
Plan sketch of beams in one quadrant for all-dielectric final optics
(upper
hemisphere)
(lower
hemisphere)
10 beams, vertical aspect ratio
set of 4 beams 11 deg to
horizontal shown here
Equatorial plane
(upper
hemisphere)
10.75m
11.85m
blanket
(lower
hemisphere)
20m
22.5m
32.5m
main containment wall
Radiation loads in all-dielectric design
HAPL Dielectric design of 4-8-06:
Plan view of one beamline
plane dielectric turning mirror, M3
focusing dielectric mirror M2, f=42m
22.5m
15.5m
20m
46cm
30cm
12.25m
10m
9.5m
1.6m
32.5m
6.0m
70cm
77cm
(access)
plane dielectric turning mirror, M1
3m
blanket
vacuum duct
main containment
(concrete)
Mirror
location
total n/FPY/cm2
M1
32.5m from target
2.3e20
M2
9.5m from M1
6e17
M3
15.5m from M2
1e16
M4
1.6m from M3
2e14
M4
6m from M3
2e13
Estimates courtesy of M. Sawan.
plane dielectric turning mirror, M4
n >0.1MeV/cm2
2.3e20
4.5e17
5e15
3e13
2e12
n>1MeV/cm2 gamma/cm2
2.3e20
8e19
3.5e17
3e17
3e15
1e16
1.5e13
2e14
1e12
2e13
Radiation resistance of multilayers compared to requirements
Category:
layer mixing
reflectivity
damage resistance
Dose:
1e19cm-2
(4% of FPY)
OK
OK (vis*)
?
1e20cm-2
(44% of FPY)
OK
anneal?
?
1e21cm-2
(4 FPY)
4nm?
anneal?
?
*I. I Orlovskiy and K. Yu. Vukolov, “Thermal and neutron tests
of multilayered dielectric mirrors” Fus. Eng. Des. 74, 865-869 (2005)
Layer mixing? (248nm mirror layers are about 35nm thick)
Mixing of layers was thought (*) to be a showstopper for
dielectric mirrors, but data on irradiated multilayers for
X-ray optics and superconductors eases this concern
(*) R. L. Bieri and M. W. Guinan, “Grazing incidence metal mirrors as the final
elements in a laser driver for inertial confinement fusion”, Fusion Technology
19 673-678 (1991).
Exposure of XUV mirrors to 1.1e19cm-2 (1-2MeV neutrons)
low index: Si or B4C, or C
high index: Mo or W
N bilayers
“d” spacing
S. P. Regan et al. “An evaluation of multilayer mirrors for the soft X ray
and extreme ultraviolet wavelength range that were irradiated with neutrons”
Rev. Sci. Instrum. 68(1), 757-760 (1997)
XUV mirrors tested by Regan et al. (exposure temp 270-300C)
Composition
“d” (nm)
substrate
N
Mo/Si
8.78nm
Zerodur
50
W/B4C
2.28nm
Si wafer
100
W/C
2.53nm
Si wafer
100
Mo/Si
18.7nm
Si wafer
25
Results:
Mo/Si mirrors had slight shift in peak L and slight reflectivity
decrease, exactly consistent with 270-300C known thermal effects.
W mirrors had opposite shift in peak L and slight increase in reflectivity,
again consistent with known thermal effects.
No effects attributable to fast neutron exposure of 1.1e19cm-2 (10-2 dpa/atom)
Layered superconductors have been tested for fusion reactor use
2nm AlN layers
9 - 26nm NbN layers (not to scale)
20 bilayers
R. Herzog et al. “Radiation effects in superconducting NbN / AlN multilayer films”
J. Appl. Phys. 68, 6327-6330 (1990).
Results: Jc>108Am-2 at 4.2K and 20T, before and after irradiation
No degradation of 2nm AlN layers at 1x1019 neutrons/cm2 (>0.1MeV)
Lack of layer mixing is consistent with expectations
XUV reflectivity and enhanced Jc are both sensitive to the roughness
of the surfaces of layers, and the above results both indicate less than
about 1nm of induced roughness after irradiation to 10-2 dpa.
Bieri and Guinan (*) estimated mixing of roughly 3nm/(dpa)1/2, where
1dpa = 5x1020 (14MeV) n/cm2 for most dielectrics
For “low” doses (much less than1 dpa), this diffusion approach does
not apply (for example, at 10-2 dpa, 99% of layer remains undisturbed).
A more critical test will occur at > 0.1dpa (5x1019 neutrons cm-2)
(20% of FPY)
(*) R. L. Bieri and M. W. Guinan, “Grazing incidence metal mirrors as the final
elements in a laser driver for inertial confinement fusion”, Fusion Technology
19 673-678 (1991).
Dielectric mirrors are being evaluated for the final optic
1.0
0.8
alumina/silica
hafnia/silica
hafnia/alumina
0.6
0.4
Reflectivity
30 layers = 1.9µm …short absorption path
Neutron-stable sapphire substrate
500C anneal cycle removes
neutron-induced absorption
0.2
0.0
10
20
Number of interfaces
30
40
Neutron irradiation produces optical absorption in dielectrics
25
Optical density in 1cm path at 257nm (4.8eV)
Al 2 O3 Abdukadyrova
Al 2 O3 Evans and Stapelbroek
20
1cm SiO 2 (248nm)
(Latkowski and Abdukadyrova)
) for 1cm path
10
Log(I
0/I
15
5
0
10
16
10
17
18
10
10
Fast neutron flux (cm
19
-2
10
)
20
10
21
Corrected transmission (%)
248nm neutron-induced absorption is annealed out at 500C
100
90
80
70
Initial
60
1 hr
50
4 days
40
2 hr @ 500C
1 day @ 500C
30
2 days @ 500C
20
3 days @ 500C
10
0
200
4 days @ 500C
300
400
500
600
700
800
Wavelength (nm)
Sample 0.6cm SiO2, 6.6e20 neutrons cm-2.
--> post anneal loss in 2µm dielectric path is negligible
Data from J. Latkowski, HAPL meeting, Rochester, Nov. 2005
Planned neutron irradiation of mirrors
All the reported multilayer exposures have been to about 1019/cm2
The key near term need is for exposure to 1020/cm2 (44% FPY)
A variety of KrF mirrors and GIMM quality aluminum mirrors will be
exposed to up to 1021 neutrons cm-2 at ORNL (>1FPY).
Some mirrors will be irradiated at high temperature (250C and 500C)
(K. J. Leonard et al. 14th laser IFE progm. wkshp, ORNL, March 2006)
GIMM design options (slide courtesy M. Sawan)
Two options considered for GIMM materials and thicknesses
Both options have 50 microns thick Al coating
Option 1: Lightweight SiC substrate
• The substrate consists of two SiC face plates surrounding a SiC foam with
12.5% density factor
• The foam is actively cooled with slow-flowing He gas
• Total thickness is 1/2"
• Total areal density is 12 kg/m2
Option 2: Lightweight AlBeMet substrate
• The substrate consists of two AlBeMet162 (62 wt.%Be) face plates
surrounding a AlBeMet foam(or honeycomb) with 12.5% density factor
• The foam is actively cooled with slow-flowing He gas
• Total thickness is 1"
• Total areal density is 16 kg/m2
Radiation loads in SiC GIMM design
5Jcm-2
Mirror
location
total n/FPY/cm2
GIMM(M1)24m from target 4.2e20
M2
14.9m from M1
1.0e18
M3
1.6m from M2
2e16
M3
6m from M2
2e15
Estimates courtesy of M. Sawan.
n >0.1MeV/cm2
4.0e20
9e17
6e15
4e14
n>1MeV/cm2 gamma/cm2
3.6e20
1.4e20
8e17
4e17
2.8e15
1.3e16
2e14
1.3e15
Unknowns remain for both the GIMM and dielectric approaches:
After neutron exposure:
Laser-induced damage has not been measured in either case
Reflectivity at 248nm (KrF) has not been measured in either case
Substrate optical quality has not been measured (except to 1019 cm-2)
Other considerations:
The GIMM requires polarized light => less random illumination
The GIMM area is 14m2 vs dielectric area 1.9m2 => cost is high
The neutron maze is much more effective in the dielectric case