Diagnostics for Fast Ignition Science
A. G. MacPhee1, K. U. Akli2, F. N. Beg3, C. D. Chen1, H. Chen1, R. Clarke4, D. S. Hey1, R. R. Freeman5,
A. J. Kemp1, M. H. Key1, J. A. King3, S. Le Pape1, A. Link5, T. Y. Ma3, H. Nakamura6, D. T. Offermann5,
V. M. Ovchinnikov5, P. K. Patel1, T. W. Phillips1, R. B. Stephens2, R. Town1, Y.Y.Tsui7, M. S. Wei3, L. D.
Van Woerkom5, A. J. Mackinnon1
1
Lawrence Livermore National Laboratory, Livermore, CA
2
General Atomics, San Diego, CA
3
Department of Mechanical and Aerospace Engineering, University of California-San Diego, La Jolla, CA
4
CCLRC Rutherford Appleton Laboratory, Oxfordshire, UK
5
College of Mathematical and Physical Sciences, The Ohio State University, Columbus, OH
6
Institute of Laser Engineering, Osaka University, Suita, Osaka, Japan
7
Department of Electrical and Computer Engineering, University of Alberta, Canada
(925) 422 3720, macphee2@llnl.gov
Abstract
The concept for Electron Fast Ignition Inertial Confinement Fusion requires sufficient
energy be transferred from the laser pulse to the assembled fuel via ~MeV electrons. We
have assembled a suite of diagnostics to characterize such transfer, simultaneously
fielding absolutely calibrated extreme ultraviolet multilayer imagers at 68 and 256eV;
spherically bent crystal imagers at 4 and 8keV; multi-keV crystal spectrometers; MeV xray bremmstrahlung, electron and proton spectrometers (along the same line of sight) and
a picosecond optical probe interferometer. These diagnostics allow careful measurement
of energy transport and deposition during and following laser-plasma interactions at
extremely high intensities in both planar and conical targets. Together with accurate onshot laser focal spot and pre-pulse characterization, these measurements are yielding new
insights into energy coupling and are providing critical data for validating numerical PIC
and hybrid PIC simulation codes in an area crucial for many applications, particularly fast
ignition. Novel aspects of these diagnostics and how they are combined to extract
quantitative data on ultra high intensity laser plasma interactions are discussed, together
with implications for full-scale fast ignition experiments.
Introduction
Fast ignition1,2 (FI) is a relatively new concept that decouples the compression and
ignition stage in inertial confinement fusion. This concept has the potential to provide a
significant advance in the technical attractiveness of Inertial Fusion Energy (IFE)
reactors. FI differs from conventional “central hot spot” (CHS) target ignition by using
one driver (laser, heavy ion beam or Z-pinch) to create a dense fuel and a separate ultrashort, ultra-intense laser beam to ignite the dense core. FI targets can burn with ~ 3X
lower density fuel than CHS targets, resulting in lower required compression energy,
relaxed drive symmetry, relaxed target smoothness tolerances, and, importantly for IFE
and ignition applications, higher gain. The short, intense ignition pulse that drives this
process interacts with extremely high energy density plasmas; the physics that controls
this interaction is only now becoming accessible in laboratory experiments through the
advent of multiple energetic short pulse lasers and the development of novel diagnostics
that allows accurate and in situ probing of energy deposition and transport in metals and
hot plasmas. Present designs3,4 for fast ignition use a hollow Au cone inserted in the
spherical shell as illustrated in Figure 1. The implosion produces the dense compressed
fuel plasma at the tip of the cone, while the hollow cone permits the short-pulse ignition
laser to be transported to the dense fuel without interference, and enables the generation
of hot electrons at its tip, very close to the dense plasma.
The critical unresolved issue is the physics of the dynamics of the 100s of MA current
propagation from the critical density near the cone tip into the assembled fuel. The
requirements on the short pulse laser have been determined from hydrodynamic and burn
calculations as described by Tabak and Atzeni et al.1,2,. These articles describe the
required energy, pulselength, heated volume size and particle range required for
successful ignition of compressed Deuterium-Tritium (DT) targets. The minimum
requirements (for compressed fuel of density 500g/cm3 and R = 2g/cm2) are hot electron
total energy Eh = 18kJ, electron beam pulselength Tp = 10ps, electron beam radius rbeam =
40m and mean particle range of 2g/cm2 (equivalent to 2MeV electrons). Assuming a
conversion efficiency from laser to (1-2MeV) electrons of 30%5,6, this results in a
required focused laser irradiance at the cone tip of 5x1020 W/cm2. According to the
ponderomotive scaling7 this laser intensity would result in an electron distribution with a
mean temperature of 5-7MeV. This is not well matched to the column density of the fuel
and results in inefficient coupling of laser to target, which subsequently increases the
energy required to ignite the target. However the validity of ponderomotive scaling has
not been adequately tested with sub-ps pulses at laser intensities above 1019W/cm2 and
there is theoretical work that suggests that hole boring by the intense laser pulse can
result in lower hot electron energy8. There is a clear need to measure both absolute
conversion efficiency and hot electron energy distribution as a function of laser intensity
in the ~1020 W/cm2 regime. Once validated, these measurements will be used to
benchmark integrated codes that will in turn be used to design integrated fast ignition (FI)
experiments1,9,10 on large scale facilities such as Omega EP, NIF ARC and Firex. The key
parameter we can study on smaller scale systems like the 150J 0.6ps Titan laser at LLNL,
is conversion efficiency from laser energy to electrons within the required bandwidth. For
these measurements to be useful to modelers, an accurate knowledge is required of the
laser pre-pulse profile, the subsequent pre-formed plasma profile, and the laser focal spot
intensity distribution. In this article we describe recent techniques designed to carry out
these measurements.
Diagnostics required for Fast Ignition experiments
1 Laser diagnostics
To accurately model the electron source produced by laser matter interactions, the
laser intensity, prepulse and resulting prefomed plasma must be characterized on every
shot. On experiments described here the laser power is monitored by measuring the onshot energy using a calorimeter and pulselength using a single shot 2nd order
autocorrelator and a grenouille11. This gives a power measurement within a shot to shot
accuracy of 20%. Figure 2 illustrates the layout of the remaining laser diagnostics. Best
focus is established at the beginning of the experiment by adjusting the parabola
alignment while monitoring the intensity distribution of the pulsed alignment beam focal
spot at the target plane using a microscope with a 16-bit CCD camera. For subsequent
system shots, the intensity distribution at the target plane is recorded with an equivalent
plane monitor (EPM). The EPM characterizes the effects of pump induced distortion and
non linear optical refractive index on the full power focal spot. The EPM records the
equivalent target plane intensity distribution from a full aperture f=6.32m f/25.28 lens
collecting 10-5 leakage through the last turning mirror prior to the final f=0.75m f/3
focusing optic. The low transmission of the final turning mirror ensures there are no B–
integral effects in the EPM system that might cause an increase in the measured focal
spot. The image is recorded on a 16-bit ccd via an infinity corrected microscope, the
collimated section of which provides a location for filters that is insensitive to their
alignment, permitting switching between low power and full system shot sensitivity
without compromising the measurement. The EPM spatial dimension scales as the focal
ratio of the two optics (6320mm/750mm = 8.43), whilst to preserve Fresnel zone number
the location of the equivalent plane scales as the square of the focal ratio
((6320mm/750mm)2 = 70.56). Figure 3 shows the intensity distribution measured with
the EPM at tight focus for both (a) a ~1mJ alignment pulse and (b) a ~150J system shot
where the scales have been normalized for clarity. Figure 4 shows the integrated power
fraction derived form the same shots, where again (a) has been normalized to (b). Clearly
the intensity distribution is consistent between the pulsed alignment beam and full system
shots. Thus it is demonstrated that 10%, 50% and 80% of available power is delivered
above 1020, 4x1019 and 1019 W/cm2 respectively. Figure 4(c) illustrates the nonlinear
relationship between power fraction and target plane location; where for a 300m
defocused system shot, 10%, 50% and 80% of the power is delivered above 1019, 5x1018
and 3x1018 W/cm2 respectively. This level of detail in our description of the intensity
distribution will be essential for providing accurate starting conditions to benchmarking
simulations.
The on-shot pre-pulse is measured on every shot using a photodiode protected
from the main pulse by a water cell. This system takes the leakage through an AR coated
wedge in the EPM that is in place for full system shots to maintain low B-integral. Hence
on system shots the majority of the leakage from the final Titan turning mirror is sent to
the photodiode. This system is capable of measuring the pulse shape up to 100ps before
the peak of the main pulse and has been used to measure the few nanosecond super
fluorescence and low energy replicas of the main pulse caused by imperfections in the
optical system. Figure 5 shows a typical trace of the laser pulse on a 150J, 0.7ps shot on
target. The signal shows the typical features that are important for determining how much
pre-plasma will be produced by this pulse. There is a slowly rising super fluorescence
that starts 2.5-3ns before the peak of main pulse, followed by a very short “spike” a few
hundred picoseconds before the main pulse. This system has shown that the Titan laser
system has an inherent prepulse of 1-10mJ, partitioned between the superfluorescence
and the spike. Both of these features are important for determining preformed plasma
conditions. (These results are described in more detail in a forthcoming publication, Y.
Tsui et al.). Together the EPM and pre-pulse monitor are essential for establishing the
correct starting conditions for the models used to characterize the electron source, as
described in the next section.
II.2 Hot electron source size
Images of the K fluorescence are used to measure the hot electron source size.
Two techniques are currently used, 2D imaging with spherically bent Bragg crystals to
measure the hot electron source size in buried fluor layers and pinhole imaging using
Ross pair12 filters and image plate detectors. Initial experiments have concentrated on
fluorescence from Ti (Z=22, K=4.5keV)13 and Cu (Z=29, K=8keV)14 fluors. However,
the relatively low ionization cross sections of Ti and Cu can push the K emission away
from the Bragg peak of the crystal, resulting in a distorted image of the hot spot and an
underestimate of the total K yield15. For this reason and to reduce the effects of opacity
for fluorescence imaging in integrated experiments, we are shifting attention to higher Z
fluors. For Zr (Z=40, K=15.7keV) we are evaluating spherically bent quartz (3140) in
3rd order16. Above Zr there are no suitable crystals for efficient 2D imaging. For Ag
(Z=47, =22keV) we are evaluating a pinhole array imager with a 30m/20m Mo/Pd
Ross pair filter (with K-edges at 20keV and 24.4keV respectively) to isolate the Ag K
signal. Figure 6 illustrates the technique using a multi-pinhole array Ross pair camera and
demonstrates a measured source size that is limited by the 30m pinhole width.
II.3 Hot electron energy distribution
The distribution of hot electrons and the associated hot electron temperature Thot is
determined from the bremsstrahlung spectrum that escapes the target. Bremsstrahlung
spectra for ultra intense laser plasma interactions extend to hundreds of keV and to
measure them we use a differential Z filter stack spectrometer17 adapted for petawatt
scale interactions18 (Figure 7a) The Monte Carlo code Integrated Tiger Series (ITS)19 is
used to construct response matrices for both the target and the detector. The ITS code
tracks electron and gamma ray showers and includes such physics as elastic and Compton
scattering, pair production, and x-ray fluorescence. For the target matrix an array of
bremsstrahlung output spectra are generated for a series of discrete electron input
energies. For the detector matrix an array of dose per dosimeter is generated for a series
of discrete X-ray energies. The electron energy distribution is varied to minimize the
error in the fit to the measured dose per channel. A one temperature Boltzmann
distribution of electrons provides a reasonable fit to the data (Figure 7b,c). For greater
precision more constraints are required for the model. Absolute calibration of the
dosimeters, both image plates and thermo luminescent detectors is underway to achieve
this and additional measurements using nuclear activation will extend the bremsstrahlung
measurement out to several MeV to further constrain the model. Figure 8 shows a one
temperature analysis of Thot scaling with laser intensity using this technique. Clearly the
electron spectrum produced in these Cu targets with 0.6ps pulses is influenced by more
than just the ponderomotive force. Work is under way to improve on the simple one
temperature fit18, and to better understand laser hole boring with sub ps pulse above
1019W/cm2 which strongly influences the plasma density profile during the interaction8.
II.4 Laser to hot electron conversion efficiency
Hot electron conversion efficiency is inferred from the K fluorescence yield induced in
the target substrate. Accurate measurements of absolute signal are obtained from crystal
spectra calibrated against an absolutely calibrated single hit CCD camera. To maintain
linearity on the CCD, it is typically necessary to maintain a single hit count rate below
~5% of the total number of pixels, with ~1-10% of these attributed to K- events. To
confirm linearity we filter one half of the single hit CCD by 0.5 Tx at the K energy and
compare the signal on both sides of the chip. Statistics for the calibration are improved by
summing over several shots. The K yields shown in Figure 9 illustrate the effect of
electron refluxing20 and laser incidence angle. The oblique incidence shots correspond to
the interaction angle used for the cone targets of similar mass, illustrating enhanced yield
from a wall confined plasma. Detailed PIC modeling is under way to correlate the K
yield with the conversion efficiency from laser energy to hot electrons. Based on current
ITS simulations that do not include ohmic potentials, existing absolute K yield data
would suggest a conversion efficiency for thick slab targets of 15%18 from laser energy to
electrons within the energy range 1-3MeV.
Conclusion
The next generation of sub-scale integrated fast ignition experiments on the OMEGA EP,
FIREX I, and NIF ARC laser facilities will require sophisticated target designs developed
with well validated modeling tools. The most important parameters related to the ignition
pulse are the mean hot electron energy and laser-to-electron conversion efficiency. Since
direct measurements of the electrons at their source are not possible one must rely on
indirect measurements, coupled with modeling to determine these quantities. We have
implemented a suite of diagnostics based on hot electron generated K-alpha x-ray and
Bremsstrahlung emission measurements, which when combined serve to constrain the
inferred hot electron distribution function and conversion efficiency. Benchmarking PIC
and hybrid PIC calculations, as well as physical scaling laws, against experimental data
additionally requires well characterized laser interaction parameters, including full energy
focal spot intensity distributions and pre-pulse measurements. The techniques described
herein using high energy x-rays to characterize the hot electron distribution can be
applied to experiments on the next generation of multi-kiloJoule class short-pulse lasers.
Such experiments will be an essential pre-requisite for the design and optimization of fast
ignition targets, and for the interpretation of measurements, such as neutron yields and Kalpha fluorescence, in integrated fast ignition experiments.
Acknowledgements
This work performed under the auspices of the U.S. Department of Energy by Lawrence
Livermore National Laboratory under Contract DE-AC52-07NA27344
Figure 1 Schematic of advanced reentrant
cone concept for fast ignition
150J 0.6ps, 1 from
pulse compressor
2, 1ps, 10mJ probe beam
Full aperture retrofocus system
Pre-pulse
monitor
`
`
6m
lens
Initial set-up
microscope
Equivalent
plane monitor
Interferometer
Figure 2 Titan chamber layout
Normalized Intensity (y) (au)
0.2
0.3
0.5
0.7
0.9
1.0
1.0
y(m)
80
0.8
0.6
60
0.4
40
0.2
Normalized Intensity (x) (au)
0.0
100
0.0
20
0
20
40
60
80
100
x(m)
Figure 3 Binned h and v profiles through EPM focus for alignment pulse (dash) and system shot (solid)
Integrated power fraction
1.0
0.8
0.6
0.4
0.2
0.0
1E17
1E18
1E19
1E20
-2
1E21
Intensity (Wcm )
Figure 4 Power fraction on target measured with EPM:
tight focus: a) system shot, b) alignment pulse (normalized
to ‘a’), c) system shot 300m upstream of best focus
ASE ~ 13 mJ
Spike ~ 8 mJ
saturated
main
-1.4 ns
pulse
-3 ns
Figure 5 Typical trace from pre-pulse monitor
0.6
Pd
Filter
0.5
Signal (psl)
Mo
Filter
Pinhole limited
width <30m
0.4
0.3
0.2
0.1
0.0
0
100
200
300
x (m)
Figure 6 Ross pair multi-pinhole array image of Ag K emission from a 12m thick Ag foil
irradiated at ~1020W/cm2
3
10
2
Signal (normalized units)
10
1
10
121 J Data
98 J Data
135 J SP, 0.8 J LP Data
Sum of Squares Deviation
121 J Data
98 J Data
135 J SP, 0.8 J LP Data
1.4 MeV
1.0 MeV
1.7 MeV
2
10
1
10
0
10
-1
10
-2
0
2
4
6
8
Dosimeter Layer
10
12
14
10 -1
10
0
10
Temperature (MeV)
Figure 7: a) Bremsstrahlung spectrometer 7b) dose and best fit to dose, 7c) T-hot fit error
1
10
2
Pondermotive scaling:
Ponderomotive scaling
T (MeV)= (I2/(1019W/cm2m2))1/2
Thot (MeV)
1.5
hot
Bremsstrahlung data
1
0.5
Beg scaling
Beg
scaling:
Thot(MeV)= 0.1(I 2/(1017W/cm2m2))1/3
0
0
50
100
150
I (1018) W/cm2
Figure 8 1 Temperature analysis of Thot scaling measured with Bremsstrahlung spectrometer
1.00E+12
20 um Cu
25 um Cu
August cones
April cones
Low mass @ 28 deg
Low mass @ 75 deg
10 Al/30 Cu August
Al/Cu/Al August
Cones have the highest yield
August cones
April cones
1.00E+11
20 um Cu
25 um Cu
multiple pass-refluxing
10 Al/30 Cu August
Low mass @ 28 deg
1.00E+10
Single pass
Al/Cu/Al August
Oblique incidence
Low mass @ 75 deg
1.00E+09
1.00E+17
1.00E+18
1.00E+19
1.00E+20
1.00E+21
Figure 9 -alpha yield (photons per steradian per incident Joule) vs peak intensity on target for a range of target configurations
1
M. Tabak, J. Hammer, M.E. Glinsky, et al, Phys. Plasmas 1, 1626 (1994).
2
S Atzeni, Phys. Plasmas 14, 052702 (2007).
3
S. Hatchett, M. Herrmann, M. Tabak, R. Turner, and R. Stephens, Bull.
Am. Phys. Soc. 46, 47 (2001)
4
S. P. Hatchett, D. Clark, M. Tabak, R. E. Turner, C. Stoeckl, R. B.
Stephens, H. Shiraga, and K. Tanaka, Fusion Sci. Technol. 49, 327 (2006).
5
K. Yasuike, M. J. Key, S. P. Hatchett, R. A. Snavely, and K. B. Wharton,
Rev. Sci. Instrum. 71, 1236 (2001).
6
K. B. Wharton, S. P. Hatchet, S. C. Wilks et al., Phys. Rev. Lett. 81, 822 (1998).
7
S. C. Wilks et al., Phys. Rev. Lett. 69, 1383 (1992).
8
A. Kemp, Y. Sentoku, M. Tabak, Submitted to Phys. Rev. Lett.
9
N.G. Basov, S.Y. Guskov, L.P. Feokistov, J. Supercrit. Fluids 13, 396 (1992).
10
M.H. Key, Phys. Plasmas, 14, 055502 (2007).
11
P. O’Shea et al., Opt. Lett. 26, 932 (2001); www.swampoptics.com
12
P.A. Ross, Phys. Rev. 28, 425 (1926)
13
J.A. King et al., Rev. Sci. Instrum. 76(7) 076102 (2005)
14
J.A. King et al., Appl. Phys. Lett 86(19) 191501 (2005)
15
K.U. Akli, Phys. Plasmas 14, 023102 (2007)
16
J.A. Koch, O.L. Landen, T.W. Barbee et al., Appl. Opt. 37 (10), 1784 (1998).
17
R. Nolte, R. Behrens, M. Schnürer, et al., Radiat. Prot. Dosim. 84 (1-4), 367 (1999)
18
C.D. Chen, these proceedings
19
J.A. Halbleib, R.P. Kensek, T.A. Mehlhorn, et al., “ITS Version 3.0: The integrated TIGER Series of
Coupled Electron/Photon Monte Carlo Transport Codes,” Technical Report SAND91-1634, Sandia
National Laboratories, 1992
20
W. Theobold, K. Akli, R. Clarke, et al., Phys. Plasmas 13, 043102 (2006)