11D608-2
Hall et al.
FIG. 1. The imaging system, showing the relative positions of the X-ray source, object, crystal, and detector.
and place a pinhole at the same position. An alignment laser can be directed into the assembly at the point shown in Fig. 3 and, using an internal mirror, illuminates the rear of the pinhole. The backlit pinhole acts as a reference point visible on the alignment camera that is used to determine the required focal spot position for the drive laser. Additionally, the target-aperture assembly ensures that there is no direct line of sight between the object and the detector, which is one of the principle advantages of backlighters based on Bragg reflection. The main body of the assembly is stainless steel, designed to block hard X-rays emitted along the direct path from the experiment to the detector. A baffle couples the film-pack chamber directly into the rear of the target-aperture assembly, ensuring that the only open path to the detector is through the aperture, and a sliding plate mounted on the front of the assembly blocks the line of sight from the aperture to the object, preventing self emission and debris from entering the aperture. These measures ensure that
Rev. Sci. Instrum. 85, 11D608 (2014)
FIG. 3. The target-aperture assembly defines the position of both the laser focal spot, and the focal spot of reflected radiation, which together determines the Bragg angle of the imaging system.
self emission from the experiment can only reach the detector via the narrow bandpass of Bragg reflection from the crystal, and since the line emission from the backlighter is usually much brighter than the object self-emission at the same wavelength, good signal to noise can be achieved even when the experiment itself is an intense emitter. The Si backlighter source was driven with a 7 J, 1 ns flattop laser pulse at 527 nm generated by frequency doubling a 20 J 1053 nm beam from a long-pulse arm of the Cerberus laser. Cerberus is a hybrid Nd:Glass and optical parametric system able to deliver a combination of ns and sub-ps probe beams to the MAGPIE target area. III. INSTRUMENT CHARACTERISATION
The resolution was measured using a 50 μm thick photochemical-machined stainless steel grid with 2 mm square holes and 0.5 mm mesh. Tungsten wires (round cross section) with diameters varying from 4 μm to 250 μm were placed across the grid as additional test objects. Figure 4(a) shows a radiograph of these objects. The mean resolution (defined as the spatial extent of a 10%-90% signal change across an edge) over the full field of view is ∼30 μm. Monochromatic backlighters with very similar imaging specifications9 have achieved resolutions of ∼10 μm, and so it is expected that the system described here is capable of significantly better than ∼30 μm resolution when fully optimized. The resolution of monochromatic backlighter systems is given by the equation:8 σ =
FIG. 2. The imaging components of the backlighter are loaded into the MAGPIE chamber on a mounting plate, whilst a separate chamber containing the film-pack is attached to a chamber port.
Sp(M + 1)(1 − sin2 θ ) , yM
(1)
where σ is the expected resolution, S is the backlighter source size, p is the object–crystal distance, y is the source–object distance, M is the magnification, and θ is the Bragg angle. Time integrated, spatially resolved spectroscopy of the backlighter source was used to measure the spatial profile of the silicon He-α line. This was found to have a FWHM of 1.2 mm, which would give σ ≈ 11 μm similar to the previously mentioned system, but also a low intensity halo
11D608-3
Hall et al.
Rev. Sci. Instrum. 85, 11D608 (2014)
FIG. 4. Radiograph of 50 μm stainless steel grid and W wires.
that extends over a ≈3.3 mm width at 10% of the peak intensity, giving σ ≈ 30 μm. This is most likely the reason for the underperformance in resolution, and the eradication of this halo is the subject for future improvements to the backlighter system. IV. INITIAL RESULTS
Recent experiments at MAGPIE include the investigation of reverse shocks relevant to astrophysical systems such as stellar jets and accretion shocks, as well as those produced by ablated material in laser driven hohlraums and Z-pinches.10 In these experiments, a reverse shock is created by placing a planar foil obstacle in the 1 × 105 m/s flow of ablated plasma produced by an inverse wire array Z-pinch. Ablated plasma stagnates on the foil, producing a reverse shock, but the electron density gradients are too large to allow measurements with laser probing diagnostics. If the ablated material is tungsten, the density of the stagnated plasma is expected to be ∼1.5 × 10−4 g/cm3 at 250 ns (assuming uniform density). Since the path length is ∼1 cm, a transmission of ∼60% might be expected for 1.865 keV X-rays (assuming cold tungsten) making the backlighter an ideal diagnostic to study conditions
within the stagnated plasma. Figure 5 shows a radiograph of a stagnated tungsten plasma and reverse shock, taken at 250 ns, and is a promising first result. The backlighter signal on this experiment was unusually low. This is partially due to the addition of an 8 μm polypropylene filter placed directly in front of the crystal to protect the surface from becoming coated with tungsten during the experiment, resulting in the loss of ∼40% of the backlighter signal. Additionally, the laser energy delivered to the target might have been reduced due to coating of the laser input window with debris from a previous experiment, and addition of protective filters to prevent this will be a future improvement to the system. V. CONCLUSIONS
A monochromatic backlighter has been developed for the MAGPIE generator, and has successfully radiographed a pulsed-power driven shock physics experiment. The diagnostic utilizes Bragg reflection from a spherically bent quartz crystal to perform radiography with 1.865 keV X-rays (Silicon He-α) driven by the Cerberus laser. This diagnostic currently achieves ∼30 μm resolution over a 4 × 20 mm field of view, although it is expected that this can be improved to ∼10 μm in the future. Additional improvements will include the upgrade of the Cerberus laser to ∼30 J, which will result in increased backlighter yield and improved signal-to-noise. ACKNOWLEDGMENTS
The author would like to acknowledge the support of the Imperial College Junior Research Fellowship scheme and the EPSRC. The Author would also like to thank Daniel B. Sinars of Sandia National Laboratories, Albuquerque, for his advice and support. 1 I.
H. Mitchell et al., Rev. Sci. Instrum. 67, 1533 (2005). A. Pikuz et al., JETP Lett. 61, 638 (1995). 3 S. A. Pikuz et al., Rev. Sci. Instrum. 68, 740 (1997). 4 D. B. Sinars et al., Rev. Sci. Instrum. 75, 3672 (2004). 5 C. Brown et al., Phys. Plasmas 4, 1397 (1997). 6 Y. Aglitskiy et al., Phys. Scr. T 132, 014021 (2008). 7 Y. Aglitskiy et al., Appl. Opt. 37, 5253 (1998). 8 D. B. Sinars et al., Appl. Opt. 42, 4059 (2003). 9 D. B. Sinars et al., Rev. Sci. Instrum. 74, 2202 (2003). 10 S. V. Lebedev et al., Phys. Plasmas 21, 056305 (2014). 2 S.
FIG. 5. Radiograph of the reverse shock produced from the stagnation of a 1 × 105 m/s laminar flow of ablated tungsten plasma onto a planar foil target on the MAGPIE generator. A low-pass signal filter and 25 × 25 μm movingmean filter have been applied to enhance features of interest.
REVIEW OF SCIENTIFIC INSTRUMENTS 85, 11D609 (2014)
Prospects for x-ray polarimetry measurements of magnetic fields in magnetized liner inertial fusion plasmasa) Alan G. Lynnb) and Mark Gilmore Department of Electrical & Computer Engineering, University of New Mexico, Albuquerque, New Mexico 87131, USA
(Presented 2 June 2014; received 31 May 2014; accepted 5 July 2014; published online 23 July 2014) Magnetized Liner Inertial Fusion (MagLIF) experiments, where a metal liner is imploded to compress a magnetized seed plasma may generate peak magnetic fields ∼104 T (100 Megagauss) over small volumes (∼10−10 m3 ) at high plasma densities (∼1028 m−3 ) on 100 ns time scales. Such conditions are extremely challenging to diagnose. We discuss the possibility of, and issues involved in, using polarimetry techniques at x-ray wavelengths to measure magnetic fields under these extreme conditions. © 2014 AIP Publishing LLC. [http://dx.doi.org/10.1063/1.4890407] I. INTRODUCTION
As high energy density plasma (HEDP) experiments advance into increasingly extreme parameter space, diagnosing plasma parameters is becoming more challenging. Plasma diagnostics commonly used in many basic laboratory plasmas and magnetic confinement systems must be significantly modified or replaced altogether in HEDP regimes. For example, polarimetry techniques at familiar far infrared and optical wavelengths, if applied to the parameters considered in this paper, would result in >600 000◦ of polarization rotation in 10’s of nanoseconds. Keeping track of this much rotation on such short timescales is not feasible, so the usual approach to polarimetry must be drastically modified. In this brief paper, we discuss the potential and challenges of extending polarimetry into the x-ray range (photon energy ∼0.1−1 keV). We will restrict our discussion here to a sample HEDP case, that of magnetized linear inertial fusion.1 II. MAGLIF EXPECTED PARAMETERS
The MagLIF concept1 involves a laser pre-heated plasma with applied axial magnetic field contained within a solid metal liner. The liner is imploded by an axial current pulse to compress the plasma to fusion conditions. We only concern ourselves with the axial magnetic field here, since the azimuthal fields that compress the liner do not penetrate into the plasma. Table I summarizes the expected parameters for a MagLIF-type plasma at early and late times during the liner compression. Here vo is plasma fluid velocity, T electron temperature, n electron number density, and Ln , Lv , LT are scale lengths of density, velocity, and temperature, respectively. The plasma axial length is given by L, with ωce electron cyclotron frequency and ωpe the electron plasma frequency. We consider a fully ionized pure deuterium plasma. a) Contributed paper, published as part of the Proceedings of the 20th
Topical Conference on High-Temperature Plasma Diagnostics, Atlanta, Georgia, USA, June 2014. b) [email protected] 0034-6748/2014/85(11)/11D609/3/$30.00
III. CONCEPTUAL OVERVIEW AND POLARIZATION ROTATION
The geometry for a conceptual polarimetry measurement based on Faraday rotation on MagLIF is shown in Fig. 1. The basic idea is to introduce a beam of polarized x-rays of appropriate photon energy through the plasma along the magnetic field B. As shown, the x-radiation does not have to be propagated exactly parallel to B, but Faraday rotation will be proportional to the k · B component, integrated over the plasma length, where k is the x-ray wavevector. After traversing the plasma, the beam is split into orthogonal linear polarizations which are sent to a pair of detectors, as shown. In the limit ω ωpe , the Faraday rotation angle can be written2 as
2 ωpe ωce B · dl ≈ 1.50 × 10−11 λ2 nLB · dl 2 (1 − ω2 /ω2 )1/2 2cω pe 0 (1) in degrees with λ x-ray wavelength, ω x-ray frequency, L plasma length traversed by the probe beam, and c speed of light. If n is known, B can be determined. Figure 2 shows the calculated Faraday rotation angle for an L = 5 mm long plasma over a range of plasma densities and axial magnetic field strengths at x-ray photon energies of 0.1 and 1 keV. It can be seen that B-fields 100T and high densities produce rotation angles >1◦ , which are relatively easily measured. Lower fields and densities, however, produce sub-degree rotation angles. Measurement of rotation angles 0.25 keV bremsstrahlung absorption is 50% loss at Ephoton < 0.1 keV. Therefore, it sets a lower limit on wavelength that is usable for measurements at peak compression. Losses from TABLE II. Sources of polarization rotation for k B.
FIG. 1. Polarized x-rays enter approximately axially along (A) where they experience Faraday rotation passing through the magnetized plasma. The xray beam then encounters a polarization separating crystal resulting in two beams with orthogonal polarizations sent along (B) and (C) to detectors.
Faraday Density gradient Velocity gradient Temperature gradient
B ≈ (Lωce /2c)(n/nc ) ∇n ≈ (sin (2θ )/16)(L/Ln )2 (n/nc )2 ∇v ≈ (vo /c)(L/Lv )(n/nc ) ∇T ≈ (T/me c2 )(L/LT )2 (n/nc )2
11D609-3
A. G. Lynn and M. Gilmore
Rev. Sci. Instrum. 85, 11D609 (2014)
TABLE III. Magnitude of polarization rotation angle in degrees using Tables I and II . We consider 0.1 keV x-rays for the rotation measurement at compression start and 1 keV x-rays at peak compression.
Ephoton ω nc B ∇n ∇v ∇T
Start of compression
Peak compression
0.1 keV 1.5 × 1017 radians/s ∼ 7 × 1030 m−3 ∼6 × 10−5 ∼ 0 (for θ ∼ 0) ∼5 × 10−11 ∼1 × 10−15
1 keV 1.5 × 1018 radians/s ∼ 7 × 1032 m−3 ∼172 ∼ 0 (for θ ∼ 0) ∼2 × 10−4 ∼6 × 10−5
VII. DETECTION SYSTEM
The simplest system to measure x-ray polarization rotation would be one based on a spectropolarimeter.12 As illustrated in Fig. 1, a crystal would configured in a specific way to split the incident x-ray beam (A) into two orthogonal polarization components traveling along paths 90◦ apart. The separate intensities would be monitored vs. time by conventional x-ray photodiodes. Another configuration, suitable for measurement of small rotation angles, could be based on interfering the probing beam with a rotating polarized reference beam,3 although such a system is more complex and would suffer lower signal-to-noise due to losses from additional xray optical components.
Compton scattering6, 7 are negligible at ≤2.3% for Ephoton ≤ 1 keV at all conditions considered. When steep gradients in refractive index are present transverse to the probe beam, refraction can become important. A detailed analysis of refraction effects would require ray tracing calculations of the probe beam, given specific n and B profiles. However, insight into the importance of refraction can be gained by calculating the ratio of the probe beam frequency, ω, to the nearest cutoff—in this case the right hand cutoff frequency2 given by ωR where ωR 2 2 1/2 = (ωce + (ωce + 4ωpe ) )/2. It can be shown that in the worst case ω/ωR ∼ 10 which suggests that refraction should not be a serious issue under the parameters considered here. Though hot plasma absorption should not be a problem, absorption by a cold solid liner will be significant. In fact, penetration of x-rays through even 100 μm of low Z liner material (e.g., Be) is near zero for photon energies less than ∼ 3 keV.6, 8
A Faraday rotation measurement of magnetic fields using x-rays in a MagLIF-type plasma would be more feasible from a purely physics standpoint at higher rather than lower magnetic fields and densities. More challenging for implementing this diagnostic would be the access required on both ends of the plasma for parallel (to B) probing, since perpendicular measurements through the solid liner making use of CottonMouton rotation do not appear to be feasible. At late times at or near stagnation, however, the solid liner remnant may be thin enough to allow ∼1 keV x-rays to penetrate. Thus, it may be possible to probe the plasma at an oblique angle, say at 30◦ from horizontal, such that k · B gives a measurable Faraday rotation angle.
VI. POLARIZED X-RAY SOURCES
The authors would like to thank M. E. Cuneo, R. D. McBride, S. B. Hansen, D. B. Sinars, and S. A. Slutz of Sandia National Laboratories for helpful discussions on MagLIF physics.
We would need x-rays of known starting polarization for a polarimetry measurement of magnetic field. This xray source will need to have sufficient intensity to provide good signal-to-noise over background x-rays and absorption from the MagLIF plasma and liner. There are a few possible x-ray sources. Synchrotrons7 and free electron lasers naturally produce polarized x-rays, but the substantial cost and infrastructure required render both impractical on a MagLIF experiment. In experiments where a high energy ultraviolet laser is available, laser-x-ray conversion from a high Z target could be used as an unpolarized source.9 If an appropriate laser is not available, a moderate size (e.g., ∼1 MA) X pinch could be used.10, 11 In either case, a Bragg crystal (either reflective or transmissive) could then be utilized to select photon energy and polarization.7 A detailed system design would need to consider the efficiency of x-ray production and losses.
VIII. CONCLUSIONS
ACKNOWLEDGMENTS
1 S.
Slutz et al., Phys. Plasmas 17, 056303 (2010). Hutchinson, Principles of Plasma Diagnostics (Cambridge University Press, 1987). 3 D. L. Brower et al., Rev. Sci. Instrum. 72, 1077 (2001). 4 T. Pisarczyk, A. Rupasov, G. Sarkisov, and A. Shikanov, J. Sov. Laser Res. 11, 1 (1990). 5 G. Bekefi, Radiation Processes in Plasmas (John Wiley & Sons, Inc., 1966). 6 Lawerence Berkeley National Laboratory Center for X-Ray Optics Database, see http://henke.lbl.gov/. 7 J. Als-Nielson and D. McMorrow, Elements of Modern X-Ray Physics, 2nd ed. (John Wiley & Sons, Ltd., 2011). 8 B. Henke, E. Gullikson, and J. Davis, At. Data Nuclear Data Tables 54, 181 (1993). 9 S. Atzeni and J. Meyer-ter-Vehn, The Physics of Inertial Fusion (Oxford University Press, 2004). 10 D. Kalantar and D. Hammer, Rev. Sci. Instrum. 66, 779 (1995). 11 D. Kalantar et al., J. Appl. Phys. 73, 8134 (1993). 12 E. Baronova, M. Stepanenko, and A. Stepanenko, Rev. Sci. Instrum. 79, 083105 (2008). 2 I.
REVIEW OF SCIENTIFIC INSTRUMENTS 85, 11D610 (2014)
Development of a ten inch manipulators-based, flexible, broadband two-crystal spectrometera) A. B. Steel,1,b) J. Dunn,1 J. Emig,1 P. Beiersdorfer,1 G. V. Brown,1 R. Shepherd,1 E. V. Marley,1 and D. J. Hoarty2 1 2
Lawrence Livermore National Laboratory, Livermore, California 94550, USA Atomic Weapons Establishment, Aldermaston, United Kingdom
(Presented 5 June 2014; received 3 June 2014; accepted 2 July 2014; published online 28 July 2014) We have developed and implemented a broadband X-ray spectrometer with a variable energy range for use at the Atomic Weapons Establishment’s Orion Laser. The spectrometer covers an energy bandwidth of ∼1–2 keV using two independently mounted, movable Bragg diffraction crystals. Using combinations of cesium hydrogen pthlate, ammonium dihydrogen phosphate, and pentaerythritol crystals, spectra covering the 1.4–2.5, 1.85–3.15, or 3.55–5.1 keV energy bands have been measured. Image plate is used for detection owing to its high dynamic range. Background signals caused by high energy X-rays and particles commonly produced in high energy laser experiments are reduced by a series of tantalum baffles and filters installed between the source and crystal and also between the crystals and detector. © 2014 AIP Publishing LLC. [http://dx.doi.org/10.1063/1.4890671] I. INTRODUCTION
A. Basic design
Bragg’s law has been used with crystals to generate x-ray spectra for over a century.1 Bragg’s law states that light incident on a crystal will be diffracted at an angle based on its angle of incidence, the lattice spacing of the crystal used, and the wavelength of the light.2 This is a well-tested technique for investigating laser-produced plasmas.3 Here, we use multiple crystals arranged side by side and offset in angle to cover a wide range of x-rays in order to characterize the plasma from which the light is emitted. The instrument was built specifically for use on buried layer experiments on the Orion laser at the Atomic Weapons Establishment (AWE), where broadband spectra give information on plasma density, temperature, and Ionization Potential Depression (IPD).4, 5
This diagnostic has the challenging task of surviving the high particle and x-ray flux and high electro-magnetic pulse (EMP) environment that is present in the Orion target chamber. LLXS itself is made of thick aluminum (a minimum of 9 mm on each side, 13 mm on the top, and 1.5 mm on the bottom). In addition to this, each side of the spectrometer is covered with 1 mm of tantalum cladding in an effort to prevent hard x-rays and hot electrons from entering. The entrance to LLXS is a removable nose cone made out of 1 mm of aluminum and has a 240 μm tantalum collimator at its end. This is the first of several baffles that act to further reduce the background counts present on the image plate. The nose cone can be replaced with an aluminum pointer for alignment. Once aligned, the image plate lies 290 mm from target chamber center, and the collimator on the front of the nose cone lies 42 mm away. Image plate was chosen as the detector because of its high dynamic range. Therefore, it is less likely to saturate from the large signal that is still present despite the filtration used.
II. EXPERIMENTAL SETUP
The Lawrence Livermore X-ray Spectrometer (LLXS) was fielded on the Orion laser at the AWE (Fig. 1). Orion consists of ten long pulse beams capable of delivering 500 J of 351 nm light with a pulse duration of 100 ps to 5 ns, and two short pulse beams delivering 500 J of 1064 nm light of durations between 0.5 and 20 ps.6 Diagnostics are mounted on Ten Inch Manipulators (TIMs) and then driven into the target chamber center to gather data. LLXS was fielded in TIM 46, which sits below target chamber center (TCC). TIM 46’s axis looks up to form an angle of 45◦ with the equatorial plane. The target was rotated such that TIM 46 was looking at the front, at an angle of 25◦ from the target surface.
a) Contributed paper, published as part of the Proceedings of the 20th
Topical Conference on High-Temperature Plasma Diagnostics, Atlanta, Georgia, USA, June 2014. b) Author to whom correspondence should be addressed. Electronic mail: [email protected]. 0034-6748/2014/85(11)/11D610/3/$30.00
B. Baffling and filtration
Along the top and sides on the inside of LLXS are 45 2.1 mm wide gaps spaced every 4 mm. These are used for removable filters and/or baffles that can be arranged for each specific crystal setup. On Orion, the first slot (closest to TCC) was filled with a filter holder that contained an 800 nm thick aluminum filter. This filter’s primary function was to act as a blast shield should any debris make it past the collimator at the front of the nose cone. In slot number two was a pair of 1 mm thick tantalum “goal post” baffles. These baffles earned this moniker by having a U shape 15.25 mm wide and 25.25 mm tall. This shape acts as coarse baffling to filter out light that would otherwise be incident on the crystals further in. In slot three was another
85, 11D610-1
© 2014 AIP Publishing LLC
11D610-2
Steel et al.
Rev. Sci. Instrum. 85, 11D610 (2014)
FIG. 1. Drawing of LLXS showing the baffle positions and crystal setup.
pair of 1 mm thick tantalum baffles. This set of baffles was laser cut for the specific crystal arrangement being used at the time and would only pass light though onto each crystal individually. The final set of baffles was also of the same design as those in slot three, but sat behind the crystals in approximately slot thirty. These baffles blocked all but the light that had diffracted off of each crystal, and, furthermore, prevented hard x-rays and hot electrons from impinging on the image plate. The total thickness of 6 mm of tantalum baffles transmits less than 10% of the photons below 7 keV, and less than 55% at 20 keV.7 The image plate itself sat in a caddy placed at the rear of the spectrometer behind a 25 μm beryllium filter. This final filter was instrumental in sifting out the low-energy rays both generated inside the spectrometer from hot electrons interacting with the aluminum body and any light that managed to pass through the baffles. C. Crystal settings
LLXS currently has baffles and crystal mounts for three different energy bands. The first setting comprises two cesium hydrogen pthlate (CsAP(001)) crystals, each 5 cm long and approximately 1.5 cm wide. The crystal mounts are made of aluminum and hold the crystals side by side, one lower than the other. Each crystal is rotated 3.5◦ from the TIM axis and sits at 3.8 and 2.5 cm below the axis. This setting covers Bragg angles from ∼11◦ to ∼20◦ . This covers an energy range of ∼1400 eV to ∼2500 eV, with a small gap of about 30 eV where the diffraction off of the two crystals does not quite overlap. Setting two is composed of one CsAP and one ammonium dihydrogen phosphate (ADP(101)) crystal. The CsAP
FIG. 2. PES target shot with a single 500 ps long-pulse beam diffracted off of an ADP crystal. Cl from the glue is also present in small quantities.
FIG. 3. Al target shot with a 500 fs shot pulse beam and diffracted off of a CsAP crystal. The self-absorption of the Al Ly-β line peak is visible at ∼2040 eV.
has a rotation of 3.35◦ from the TIM axis and sits 2.5 cm below it. This allows a limited range of ∼1850 eV to ∼2175 eV. The ADP crystal is rotated 10◦ from the TIM axis and sits 4.75 cm below it. Good data were produced despite the inherent low reflectivity of the ADP crystal. It produced spectra in the range of ∼2750 eV to ∼3180 eV. Therefore, this setting has a rather significant gap between the crystals of ∼600 eV. The third setting uses two pentaerythritol (PET(002)) crystals, each the same size as the previous CsAP and ADP crystals. PET’s small 2d spacing relative to CsAP allows the crystals to cover much higher energies with a higher spectral resolution. The crystals are rotated at angles of 6◦ and 4.5◦ from the TIM axis and sit 4.9 and 4 cm below it, respectively. This gives an energy coverage of ∼3550 eV to ∼5250 eV while producing an overlap between the two spectra substantial enough to combine the image plate lineouts into one spectra. III. RESULTS
Results from measurements on Orion are shown in Figures 2–7. These were obtained using settings two and three. For each set of data shown, the filter response has been removed and the dispersion of each crystal calculated based on an observation of K-shell transitions from hydrogen-like and helium-like ions with known energies. In order to calculate the dispersion for the second setting, two different targets were needed, one for each crystal. The ADP crystal’s coverage was measured using a 50 μm thick polyethersulphone
FIG. 4. A PET crystal was used to investigate a NaCl target shot with two 800 ps long pulse beams. Only the lower energy crystal from this setting is shown, as the higher energy crystal contained only continuum and crystal artifacts.
11D610-3
Steel et al.
Rev. Sci. Instrum. 85, 11D610 (2014)
FIG. 5. Image plate from KCl target. The two PET crystals have been moved to show the overlap between them.
(PES, OC6 H4 OC6 H4 SO2 C6 H4 ) target. This plastic was attached to the target holder with glue that contained Cl. This caused some spectral contamination, and was particularly troublesome on the few shots where almost no sulphur was present. Figure 2 shows a good example of the PES targets. The CsAP crystal dispersion on setting two was calculated using a 200 nm thick Al foil buried under 5 μm of CH. This target was shot using a 500 fs short pulse beam; therefore the densities were much higher (and the line widths much broader) than on the other spectra. Figure 2 shows that LLXS was able to spectrally resolve the self-absorption of the Al Ly-β line peak. The third setting’s dispersion was fit using two types of tamped targets, KCl and NaCl. These targets were each shot with two 800 ps pulses, timed such that when the first was ending the second began. This allowed the targets to expand into the vacuum and blow down to lower densities. The low densities prevented continuum lowering, allowing high n-shell energy states to be present. NaCl was used to help distinguish what lines originated from the K and what lines were formed from the Cl. Figure 4 shows an example from 790 nm of NaCl buried under 310 nm of Parylene-N. Both the Cl He- and Ly-series limits are visible down to the zeta (n = 7 to n = 1 transition) lines. Once the Cl lines were identified, they were used to sort through the KCl spectra. The K lines were significantly brighter than the Cl lines, as shown in Figures 5–7. Each crystal on setting three was fit on its own, and then combined for the final spectra. In Figures 6 and 7 the lower energy crystal is shown in blue, the higher energy crystal in purple. Three line features overlap, and their intensities agree very well with no normalization needed. More detail can be seen in Figure 7, all four (Cl He-, Cl Ly-, K He-, and K Ly-) series limits are clearly visible.
FIG. 7. Zoomed in KCl spectra. The Cl He-series, Cl Ly-series, K He-series, and K Ly-series limits are all clearly visible down to the zeta lines.
IV. DISCUSSION
Both tested settings were shown to produce quality spectra for long pulse beam shots. Short pulse data were also collected; however, the data were not as clean with the exception of the Al data shown. This could be resolved with more filtration to better sort out the hard x-rays produced during high intensity laser shots. The data that were gathered by LLXS during this experimental campaign can be used to determine the temperature and density of the plasma by measuring spectral line ratios and line widths, respectively. The data produced from setting three may also have a wide enough spectral range to determine the plasma temperature from the free-bound continuum as well. This gives a self-consistent means of checking the initial measurement without having to compare the result with a second diagnostic. Each type of crystal used (CsAP, ADP, and PET) produced fine spectra. The data from the ADP crystal are of particular note, as ADP has approximately 5 times lower reflectivity than the other crystals tested for their respective energy ranges.8 The LLXS design can accommodate with a CCD detector sitting ∼30 cm from TCC, therefore it is attached to an airbox with an onboard computer and cooling system. However, after a few days of shots the thermoelectric cooler on the CCD malfunctioned. It is currently hypothesized that the EMP generated from the high flux environment of Orion was too great for the current shielding on the CCD. Additional shielding will be implemented on further campaigns. ACKNOWLEDGMENTS
This work performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract No. DE-AC52-07NA27344. 1 W.
L. Bragg, Proc. Camb. Philos. Soc. 17, 43 (1912).
2 A. H. Compton and S. K. Allison, X-ray in Theory and Experiment (D. Van
Nostrand Company, New York, 1935). U. Akli et al., J. Instrum. 5, P07008 (2010). 4 D. J. Hoarty et al., High Energy Density Phys. 9, 661 (2013). 5 D. J. Hoarty et al., Phys. Rev. Lett. 110, 265003 (2013). 6 N. Hopps et al., “Overview of project Orion,” in Proceedings of the SPIE 7916, High Power Lasers for Fusion Research, 79160C, 18 February, 2011. 7 B. L. Henke, E. M. Gullikson, and J. C. Davis, “X-ray interactions: Photoabsorption, scattering, transmission, and reflection at E = 50–30000 eV, Z = 1–92,” Atomic Data Nucl. Data Tables 54(2), 181–342 (1993). 8 A. J. Burek, Space Sci. Instrum. 2, 53–104 (1976). 3 K.
FIG. 6. A target of 790 nm KCl sandwiched between two layers 310 nm of Parylene-N was shot with two 800 ps long pulse beams. Spectra from both PET crystals were combined to form one continuous spectrum.
Review of Scientific Instruments is copyrighted by the American Institute of Physics (AIP). Redistribution of journal material is subject to the AIP online journal license and/or AIP copyright. For more information, see http://ojps.aip.org/rsio/rsicr.jsp
Hall et al.
FIG. 1. The imaging system, showing the relative positions of the X-ray source, object, crystal, and detector.
and place a pinhole at the same position. An alignment laser can be directed into the assembly at the point shown in Fig. 3 and, using an internal mirror, illuminates the rear of the pinhole. The backlit pinhole acts as a reference point visible on the alignment camera that is used to determine the required focal spot position for the drive laser. Additionally, the target-aperture assembly ensures that there is no direct line of sight between the object and the detector, which is one of the principle advantages of backlighters based on Bragg reflection. The main body of the assembly is stainless steel, designed to block hard X-rays emitted along the direct path from the experiment to the detector. A baffle couples the film-pack chamber directly into the rear of the target-aperture assembly, ensuring that the only open path to the detector is through the aperture, and a sliding plate mounted on the front of the assembly blocks the line of sight from the aperture to the object, preventing self emission and debris from entering the aperture. These measures ensure that
Rev. Sci. Instrum. 85, 11D608 (2014)
FIG. 3. The target-aperture assembly defines the position of both the laser focal spot, and the focal spot of reflected radiation, which together determines the Bragg angle of the imaging system.
self emission from the experiment can only reach the detector via the narrow bandpass of Bragg reflection from the crystal, and since the line emission from the backlighter is usually much brighter than the object self-emission at the same wavelength, good signal to noise can be achieved even when the experiment itself is an intense emitter. The Si backlighter source was driven with a 7 J, 1 ns flattop laser pulse at 527 nm generated by frequency doubling a 20 J 1053 nm beam from a long-pulse arm of the Cerberus laser. Cerberus is a hybrid Nd:Glass and optical parametric system able to deliver a combination of ns and sub-ps probe beams to the MAGPIE target area. III. INSTRUMENT CHARACTERISATION
The resolution was measured using a 50 μm thick photochemical-machined stainless steel grid with 2 mm square holes and 0.5 mm mesh. Tungsten wires (round cross section) with diameters varying from 4 μm to 250 μm were placed across the grid as additional test objects. Figure 4(a) shows a radiograph of these objects. The mean resolution (defined as the spatial extent of a 10%-90% signal change across an edge) over the full field of view is ∼30 μm. Monochromatic backlighters with very similar imaging specifications9 have achieved resolutions of ∼10 μm, and so it is expected that the system described here is capable of significantly better than ∼30 μm resolution when fully optimized. The resolution of monochromatic backlighter systems is given by the equation:8 σ =
FIG. 2. The imaging components of the backlighter are loaded into the MAGPIE chamber on a mounting plate, whilst a separate chamber containing the film-pack is attached to a chamber port.
Sp(M + 1)(1 − sin2 θ ) , yM
(1)
where σ is the expected resolution, S is the backlighter source size, p is the object–crystal distance, y is the source–object distance, M is the magnification, and θ is the Bragg angle. Time integrated, spatially resolved spectroscopy of the backlighter source was used to measure the spatial profile of the silicon He-α line. This was found to have a FWHM of 1.2 mm, which would give σ ≈ 11 μm similar to the previously mentioned system, but also a low intensity halo
11D608-3
Hall et al.
Rev. Sci. Instrum. 85, 11D608 (2014)
FIG. 4. Radiograph of 50 μm stainless steel grid and W wires.
that extends over a ≈3.3 mm width at 10% of the peak intensity, giving σ ≈ 30 μm. This is most likely the reason for the underperformance in resolution, and the eradication of this halo is the subject for future improvements to the backlighter system. IV. INITIAL RESULTS
Recent experiments at MAGPIE include the investigation of reverse shocks relevant to astrophysical systems such as stellar jets and accretion shocks, as well as those produced by ablated material in laser driven hohlraums and Z-pinches.10 In these experiments, a reverse shock is created by placing a planar foil obstacle in the 1 × 105 m/s flow of ablated plasma produced by an inverse wire array Z-pinch. Ablated plasma stagnates on the foil, producing a reverse shock, but the electron density gradients are too large to allow measurements with laser probing diagnostics. If the ablated material is tungsten, the density of the stagnated plasma is expected to be ∼1.5 × 10−4 g/cm3 at 250 ns (assuming uniform density). Since the path length is ∼1 cm, a transmission of ∼60% might be expected for 1.865 keV X-rays (assuming cold tungsten) making the backlighter an ideal diagnostic to study conditions
within the stagnated plasma. Figure 5 shows a radiograph of a stagnated tungsten plasma and reverse shock, taken at 250 ns, and is a promising first result. The backlighter signal on this experiment was unusually low. This is partially due to the addition of an 8 μm polypropylene filter placed directly in front of the crystal to protect the surface from becoming coated with tungsten during the experiment, resulting in the loss of ∼40% of the backlighter signal. Additionally, the laser energy delivered to the target might have been reduced due to coating of the laser input window with debris from a previous experiment, and addition of protective filters to prevent this will be a future improvement to the system. V. CONCLUSIONS
A monochromatic backlighter has been developed for the MAGPIE generator, and has successfully radiographed a pulsed-power driven shock physics experiment. The diagnostic utilizes Bragg reflection from a spherically bent quartz crystal to perform radiography with 1.865 keV X-rays (Silicon He-α) driven by the Cerberus laser. This diagnostic currently achieves ∼30 μm resolution over a 4 × 20 mm field of view, although it is expected that this can be improved to ∼10 μm in the future. Additional improvements will include the upgrade of the Cerberus laser to ∼30 J, which will result in increased backlighter yield and improved signal-to-noise. ACKNOWLEDGMENTS
The author would like to acknowledge the support of the Imperial College Junior Research Fellowship scheme and the EPSRC. The Author would also like to thank Daniel B. Sinars of Sandia National Laboratories, Albuquerque, for his advice and support. 1 I.
H. Mitchell et al., Rev. Sci. Instrum. 67, 1533 (2005). A. Pikuz et al., JETP Lett. 61, 638 (1995). 3 S. A. Pikuz et al., Rev. Sci. Instrum. 68, 740 (1997). 4 D. B. Sinars et al., Rev. Sci. Instrum. 75, 3672 (2004). 5 C. Brown et al., Phys. Plasmas 4, 1397 (1997). 6 Y. Aglitskiy et al., Phys. Scr. T 132, 014021 (2008). 7 Y. Aglitskiy et al., Appl. Opt. 37, 5253 (1998). 8 D. B. Sinars et al., Appl. Opt. 42, 4059 (2003). 9 D. B. Sinars et al., Rev. Sci. Instrum. 74, 2202 (2003). 10 S. V. Lebedev et al., Phys. Plasmas 21, 056305 (2014). 2 S.
FIG. 5. Radiograph of the reverse shock produced from the stagnation of a 1 × 105 m/s laminar flow of ablated tungsten plasma onto a planar foil target on the MAGPIE generator. A low-pass signal filter and 25 × 25 μm movingmean filter have been applied to enhance features of interest.
REVIEW OF SCIENTIFIC INSTRUMENTS 85, 11D609 (2014)
Prospects for x-ray polarimetry measurements of magnetic fields in magnetized liner inertial fusion plasmasa) Alan G. Lynnb) and Mark Gilmore Department of Electrical & Computer Engineering, University of New Mexico, Albuquerque, New Mexico 87131, USA
(Presented 2 June 2014; received 31 May 2014; accepted 5 July 2014; published online 23 July 2014) Magnetized Liner Inertial Fusion (MagLIF) experiments, where a metal liner is imploded to compress a magnetized seed plasma may generate peak magnetic fields ∼104 T (100 Megagauss) over small volumes (∼10−10 m3 ) at high plasma densities (∼1028 m−3 ) on 100 ns time scales. Such conditions are extremely challenging to diagnose. We discuss the possibility of, and issues involved in, using polarimetry techniques at x-ray wavelengths to measure magnetic fields under these extreme conditions. © 2014 AIP Publishing LLC. [http://dx.doi.org/10.1063/1.4890407] I. INTRODUCTION
As high energy density plasma (HEDP) experiments advance into increasingly extreme parameter space, diagnosing plasma parameters is becoming more challenging. Plasma diagnostics commonly used in many basic laboratory plasmas and magnetic confinement systems must be significantly modified or replaced altogether in HEDP regimes. For example, polarimetry techniques at familiar far infrared and optical wavelengths, if applied to the parameters considered in this paper, would result in >600 000◦ of polarization rotation in 10’s of nanoseconds. Keeping track of this much rotation on such short timescales is not feasible, so the usual approach to polarimetry must be drastically modified. In this brief paper, we discuss the potential and challenges of extending polarimetry into the x-ray range (photon energy ∼0.1−1 keV). We will restrict our discussion here to a sample HEDP case, that of magnetized linear inertial fusion.1 II. MAGLIF EXPECTED PARAMETERS
The MagLIF concept1 involves a laser pre-heated plasma with applied axial magnetic field contained within a solid metal liner. The liner is imploded by an axial current pulse to compress the plasma to fusion conditions. We only concern ourselves with the axial magnetic field here, since the azimuthal fields that compress the liner do not penetrate into the plasma. Table I summarizes the expected parameters for a MagLIF-type plasma at early and late times during the liner compression. Here vo is plasma fluid velocity, T electron temperature, n electron number density, and Ln , Lv , LT are scale lengths of density, velocity, and temperature, respectively. The plasma axial length is given by L, with ωce electron cyclotron frequency and ωpe the electron plasma frequency. We consider a fully ionized pure deuterium plasma. a) Contributed paper, published as part of the Proceedings of the 20th
Topical Conference on High-Temperature Plasma Diagnostics, Atlanta, Georgia, USA, June 2014. b) [email protected] 0034-6748/2014/85(11)/11D609/3/$30.00
III. CONCEPTUAL OVERVIEW AND POLARIZATION ROTATION
The geometry for a conceptual polarimetry measurement based on Faraday rotation on MagLIF is shown in Fig. 1. The basic idea is to introduce a beam of polarized x-rays of appropriate photon energy through the plasma along the magnetic field B. As shown, the x-radiation does not have to be propagated exactly parallel to B, but Faraday rotation will be proportional to the k · B component, integrated over the plasma length, where k is the x-ray wavevector. After traversing the plasma, the beam is split into orthogonal linear polarizations which are sent to a pair of detectors, as shown. In the limit ω ωpe , the Faraday rotation angle can be written2 as
2 ωpe ωce B · dl ≈ 1.50 × 10−11 λ2 nLB · dl 2 (1 − ω2 /ω2 )1/2 2cω pe 0 (1) in degrees with λ x-ray wavelength, ω x-ray frequency, L plasma length traversed by the probe beam, and c speed of light. If n is known, B can be determined. Figure 2 shows the calculated Faraday rotation angle for an L = 5 mm long plasma over a range of plasma densities and axial magnetic field strengths at x-ray photon energies of 0.1 and 1 keV. It can be seen that B-fields 100T and high densities produce rotation angles >1◦ , which are relatively easily measured. Lower fields and densities, however, produce sub-degree rotation angles. Measurement of rotation angles 0.25 keV bremsstrahlung absorption is 50% loss at Ephoton < 0.1 keV. Therefore, it sets a lower limit on wavelength that is usable for measurements at peak compression. Losses from TABLE II. Sources of polarization rotation for k B.
FIG. 1. Polarized x-rays enter approximately axially along (A) where they experience Faraday rotation passing through the magnetized plasma. The xray beam then encounters a polarization separating crystal resulting in two beams with orthogonal polarizations sent along (B) and (C) to detectors.
Faraday Density gradient Velocity gradient Temperature gradient
B ≈ (Lωce /2c)(n/nc ) ∇n ≈ (sin (2θ )/16)(L/Ln )2 (n/nc )2 ∇v ≈ (vo /c)(L/Lv )(n/nc ) ∇T ≈ (T/me c2 )(L/LT )2 (n/nc )2
11D609-3
A. G. Lynn and M. Gilmore
Rev. Sci. Instrum. 85, 11D609 (2014)
TABLE III. Magnitude of polarization rotation angle in degrees using Tables I and II . We consider 0.1 keV x-rays for the rotation measurement at compression start and 1 keV x-rays at peak compression.
Ephoton ω nc B ∇n ∇v ∇T
Start of compression
Peak compression
0.1 keV 1.5 × 1017 radians/s ∼ 7 × 1030 m−3 ∼6 × 10−5 ∼ 0 (for θ ∼ 0) ∼5 × 10−11 ∼1 × 10−15
1 keV 1.5 × 1018 radians/s ∼ 7 × 1032 m−3 ∼172 ∼ 0 (for θ ∼ 0) ∼2 × 10−4 ∼6 × 10−5
VII. DETECTION SYSTEM
The simplest system to measure x-ray polarization rotation would be one based on a spectropolarimeter.12 As illustrated in Fig. 1, a crystal would configured in a specific way to split the incident x-ray beam (A) into two orthogonal polarization components traveling along paths 90◦ apart. The separate intensities would be monitored vs. time by conventional x-ray photodiodes. Another configuration, suitable for measurement of small rotation angles, could be based on interfering the probing beam with a rotating polarized reference beam,3 although such a system is more complex and would suffer lower signal-to-noise due to losses from additional xray optical components.
Compton scattering6, 7 are negligible at ≤2.3% for Ephoton ≤ 1 keV at all conditions considered. When steep gradients in refractive index are present transverse to the probe beam, refraction can become important. A detailed analysis of refraction effects would require ray tracing calculations of the probe beam, given specific n and B profiles. However, insight into the importance of refraction can be gained by calculating the ratio of the probe beam frequency, ω, to the nearest cutoff—in this case the right hand cutoff frequency2 given by ωR where ωR 2 2 1/2 = (ωce + (ωce + 4ωpe ) )/2. It can be shown that in the worst case ω/ωR ∼ 10 which suggests that refraction should not be a serious issue under the parameters considered here. Though hot plasma absorption should not be a problem, absorption by a cold solid liner will be significant. In fact, penetration of x-rays through even 100 μm of low Z liner material (e.g., Be) is near zero for photon energies less than ∼ 3 keV.6, 8
A Faraday rotation measurement of magnetic fields using x-rays in a MagLIF-type plasma would be more feasible from a purely physics standpoint at higher rather than lower magnetic fields and densities. More challenging for implementing this diagnostic would be the access required on both ends of the plasma for parallel (to B) probing, since perpendicular measurements through the solid liner making use of CottonMouton rotation do not appear to be feasible. At late times at or near stagnation, however, the solid liner remnant may be thin enough to allow ∼1 keV x-rays to penetrate. Thus, it may be possible to probe the plasma at an oblique angle, say at 30◦ from horizontal, such that k · B gives a measurable Faraday rotation angle.
VI. POLARIZED X-RAY SOURCES
The authors would like to thank M. E. Cuneo, R. D. McBride, S. B. Hansen, D. B. Sinars, and S. A. Slutz of Sandia National Laboratories for helpful discussions on MagLIF physics.
We would need x-rays of known starting polarization for a polarimetry measurement of magnetic field. This xray source will need to have sufficient intensity to provide good signal-to-noise over background x-rays and absorption from the MagLIF plasma and liner. There are a few possible x-ray sources. Synchrotrons7 and free electron lasers naturally produce polarized x-rays, but the substantial cost and infrastructure required render both impractical on a MagLIF experiment. In experiments where a high energy ultraviolet laser is available, laser-x-ray conversion from a high Z target could be used as an unpolarized source.9 If an appropriate laser is not available, a moderate size (e.g., ∼1 MA) X pinch could be used.10, 11 In either case, a Bragg crystal (either reflective or transmissive) could then be utilized to select photon energy and polarization.7 A detailed system design would need to consider the efficiency of x-ray production and losses.
VIII. CONCLUSIONS
ACKNOWLEDGMENTS
1 S.
Slutz et al., Phys. Plasmas 17, 056303 (2010). Hutchinson, Principles of Plasma Diagnostics (Cambridge University Press, 1987). 3 D. L. Brower et al., Rev. Sci. Instrum. 72, 1077 (2001). 4 T. Pisarczyk, A. Rupasov, G. Sarkisov, and A. Shikanov, J. Sov. Laser Res. 11, 1 (1990). 5 G. Bekefi, Radiation Processes in Plasmas (John Wiley & Sons, Inc., 1966). 6 Lawerence Berkeley National Laboratory Center for X-Ray Optics Database, see http://henke.lbl.gov/. 7 J. Als-Nielson and D. McMorrow, Elements of Modern X-Ray Physics, 2nd ed. (John Wiley & Sons, Ltd., 2011). 8 B. Henke, E. Gullikson, and J. Davis, At. Data Nuclear Data Tables 54, 181 (1993). 9 S. Atzeni and J. Meyer-ter-Vehn, The Physics of Inertial Fusion (Oxford University Press, 2004). 10 D. Kalantar and D. Hammer, Rev. Sci. Instrum. 66, 779 (1995). 11 D. Kalantar et al., J. Appl. Phys. 73, 8134 (1993). 12 E. Baronova, M. Stepanenko, and A. Stepanenko, Rev. Sci. Instrum. 79, 083105 (2008). 2 I.
REVIEW OF SCIENTIFIC INSTRUMENTS 85, 11D610 (2014)
Development of a ten inch manipulators-based, flexible, broadband two-crystal spectrometera) A. B. Steel,1,b) J. Dunn,1 J. Emig,1 P. Beiersdorfer,1 G. V. Brown,1 R. Shepherd,1 E. V. Marley,1 and D. J. Hoarty2 1 2
Lawrence Livermore National Laboratory, Livermore, California 94550, USA Atomic Weapons Establishment, Aldermaston, United Kingdom
(Presented 5 June 2014; received 3 June 2014; accepted 2 July 2014; published online 28 July 2014) We have developed and implemented a broadband X-ray spectrometer with a variable energy range for use at the Atomic Weapons Establishment’s Orion Laser. The spectrometer covers an energy bandwidth of ∼1–2 keV using two independently mounted, movable Bragg diffraction crystals. Using combinations of cesium hydrogen pthlate, ammonium dihydrogen phosphate, and pentaerythritol crystals, spectra covering the 1.4–2.5, 1.85–3.15, or 3.55–5.1 keV energy bands have been measured. Image plate is used for detection owing to its high dynamic range. Background signals caused by high energy X-rays and particles commonly produced in high energy laser experiments are reduced by a series of tantalum baffles and filters installed between the source and crystal and also between the crystals and detector. © 2014 AIP Publishing LLC. [http://dx.doi.org/10.1063/1.4890671] I. INTRODUCTION
A. Basic design
Bragg’s law has been used with crystals to generate x-ray spectra for over a century.1 Bragg’s law states that light incident on a crystal will be diffracted at an angle based on its angle of incidence, the lattice spacing of the crystal used, and the wavelength of the light.2 This is a well-tested technique for investigating laser-produced plasmas.3 Here, we use multiple crystals arranged side by side and offset in angle to cover a wide range of x-rays in order to characterize the plasma from which the light is emitted. The instrument was built specifically for use on buried layer experiments on the Orion laser at the Atomic Weapons Establishment (AWE), where broadband spectra give information on plasma density, temperature, and Ionization Potential Depression (IPD).4, 5
This diagnostic has the challenging task of surviving the high particle and x-ray flux and high electro-magnetic pulse (EMP) environment that is present in the Orion target chamber. LLXS itself is made of thick aluminum (a minimum of 9 mm on each side, 13 mm on the top, and 1.5 mm on the bottom). In addition to this, each side of the spectrometer is covered with 1 mm of tantalum cladding in an effort to prevent hard x-rays and hot electrons from entering. The entrance to LLXS is a removable nose cone made out of 1 mm of aluminum and has a 240 μm tantalum collimator at its end. This is the first of several baffles that act to further reduce the background counts present on the image plate. The nose cone can be replaced with an aluminum pointer for alignment. Once aligned, the image plate lies 290 mm from target chamber center, and the collimator on the front of the nose cone lies 42 mm away. Image plate was chosen as the detector because of its high dynamic range. Therefore, it is less likely to saturate from the large signal that is still present despite the filtration used.
II. EXPERIMENTAL SETUP
The Lawrence Livermore X-ray Spectrometer (LLXS) was fielded on the Orion laser at the AWE (Fig. 1). Orion consists of ten long pulse beams capable of delivering 500 J of 351 nm light with a pulse duration of 100 ps to 5 ns, and two short pulse beams delivering 500 J of 1064 nm light of durations between 0.5 and 20 ps.6 Diagnostics are mounted on Ten Inch Manipulators (TIMs) and then driven into the target chamber center to gather data. LLXS was fielded in TIM 46, which sits below target chamber center (TCC). TIM 46’s axis looks up to form an angle of 45◦ with the equatorial plane. The target was rotated such that TIM 46 was looking at the front, at an angle of 25◦ from the target surface.
a) Contributed paper, published as part of the Proceedings of the 20th
Topical Conference on High-Temperature Plasma Diagnostics, Atlanta, Georgia, USA, June 2014. b) Author to whom correspondence should be addressed. Electronic mail: [email protected]. 0034-6748/2014/85(11)/11D610/3/$30.00
B. Baffling and filtration
Along the top and sides on the inside of LLXS are 45 2.1 mm wide gaps spaced every 4 mm. These are used for removable filters and/or baffles that can be arranged for each specific crystal setup. On Orion, the first slot (closest to TCC) was filled with a filter holder that contained an 800 nm thick aluminum filter. This filter’s primary function was to act as a blast shield should any debris make it past the collimator at the front of the nose cone. In slot number two was a pair of 1 mm thick tantalum “goal post” baffles. These baffles earned this moniker by having a U shape 15.25 mm wide and 25.25 mm tall. This shape acts as coarse baffling to filter out light that would otherwise be incident on the crystals further in. In slot three was another
85, 11D610-1
© 2014 AIP Publishing LLC
11D610-2
Steel et al.
Rev. Sci. Instrum. 85, 11D610 (2014)
FIG. 1. Drawing of LLXS showing the baffle positions and crystal setup.
pair of 1 mm thick tantalum baffles. This set of baffles was laser cut for the specific crystal arrangement being used at the time and would only pass light though onto each crystal individually. The final set of baffles was also of the same design as those in slot three, but sat behind the crystals in approximately slot thirty. These baffles blocked all but the light that had diffracted off of each crystal, and, furthermore, prevented hard x-rays and hot electrons from impinging on the image plate. The total thickness of 6 mm of tantalum baffles transmits less than 10% of the photons below 7 keV, and less than 55% at 20 keV.7 The image plate itself sat in a caddy placed at the rear of the spectrometer behind a 25 μm beryllium filter. This final filter was instrumental in sifting out the low-energy rays both generated inside the spectrometer from hot electrons interacting with the aluminum body and any light that managed to pass through the baffles. C. Crystal settings
LLXS currently has baffles and crystal mounts for three different energy bands. The first setting comprises two cesium hydrogen pthlate (CsAP(001)) crystals, each 5 cm long and approximately 1.5 cm wide. The crystal mounts are made of aluminum and hold the crystals side by side, one lower than the other. Each crystal is rotated 3.5◦ from the TIM axis and sits at 3.8 and 2.5 cm below the axis. This setting covers Bragg angles from ∼11◦ to ∼20◦ . This covers an energy range of ∼1400 eV to ∼2500 eV, with a small gap of about 30 eV where the diffraction off of the two crystals does not quite overlap. Setting two is composed of one CsAP and one ammonium dihydrogen phosphate (ADP(101)) crystal. The CsAP
FIG. 2. PES target shot with a single 500 ps long-pulse beam diffracted off of an ADP crystal. Cl from the glue is also present in small quantities.
FIG. 3. Al target shot with a 500 fs shot pulse beam and diffracted off of a CsAP crystal. The self-absorption of the Al Ly-β line peak is visible at ∼2040 eV.
has a rotation of 3.35◦ from the TIM axis and sits 2.5 cm below it. This allows a limited range of ∼1850 eV to ∼2175 eV. The ADP crystal is rotated 10◦ from the TIM axis and sits 4.75 cm below it. Good data were produced despite the inherent low reflectivity of the ADP crystal. It produced spectra in the range of ∼2750 eV to ∼3180 eV. Therefore, this setting has a rather significant gap between the crystals of ∼600 eV. The third setting uses two pentaerythritol (PET(002)) crystals, each the same size as the previous CsAP and ADP crystals. PET’s small 2d spacing relative to CsAP allows the crystals to cover much higher energies with a higher spectral resolution. The crystals are rotated at angles of 6◦ and 4.5◦ from the TIM axis and sit 4.9 and 4 cm below it, respectively. This gives an energy coverage of ∼3550 eV to ∼5250 eV while producing an overlap between the two spectra substantial enough to combine the image plate lineouts into one spectra. III. RESULTS
Results from measurements on Orion are shown in Figures 2–7. These were obtained using settings two and three. For each set of data shown, the filter response has been removed and the dispersion of each crystal calculated based on an observation of K-shell transitions from hydrogen-like and helium-like ions with known energies. In order to calculate the dispersion for the second setting, two different targets were needed, one for each crystal. The ADP crystal’s coverage was measured using a 50 μm thick polyethersulphone
FIG. 4. A PET crystal was used to investigate a NaCl target shot with two 800 ps long pulse beams. Only the lower energy crystal from this setting is shown, as the higher energy crystal contained only continuum and crystal artifacts.
11D610-3
Steel et al.
Rev. Sci. Instrum. 85, 11D610 (2014)
FIG. 5. Image plate from KCl target. The two PET crystals have been moved to show the overlap between them.
(PES, OC6 H4 OC6 H4 SO2 C6 H4 ) target. This plastic was attached to the target holder with glue that contained Cl. This caused some spectral contamination, and was particularly troublesome on the few shots where almost no sulphur was present. Figure 2 shows a good example of the PES targets. The CsAP crystal dispersion on setting two was calculated using a 200 nm thick Al foil buried under 5 μm of CH. This target was shot using a 500 fs short pulse beam; therefore the densities were much higher (and the line widths much broader) than on the other spectra. Figure 2 shows that LLXS was able to spectrally resolve the self-absorption of the Al Ly-β line peak. The third setting’s dispersion was fit using two types of tamped targets, KCl and NaCl. These targets were each shot with two 800 ps pulses, timed such that when the first was ending the second began. This allowed the targets to expand into the vacuum and blow down to lower densities. The low densities prevented continuum lowering, allowing high n-shell energy states to be present. NaCl was used to help distinguish what lines originated from the K and what lines were formed from the Cl. Figure 4 shows an example from 790 nm of NaCl buried under 310 nm of Parylene-N. Both the Cl He- and Ly-series limits are visible down to the zeta (n = 7 to n = 1 transition) lines. Once the Cl lines were identified, they were used to sort through the KCl spectra. The K lines were significantly brighter than the Cl lines, as shown in Figures 5–7. Each crystal on setting three was fit on its own, and then combined for the final spectra. In Figures 6 and 7 the lower energy crystal is shown in blue, the higher energy crystal in purple. Three line features overlap, and their intensities agree very well with no normalization needed. More detail can be seen in Figure 7, all four (Cl He-, Cl Ly-, K He-, and K Ly-) series limits are clearly visible.
FIG. 7. Zoomed in KCl spectra. The Cl He-series, Cl Ly-series, K He-series, and K Ly-series limits are all clearly visible down to the zeta lines.
IV. DISCUSSION
Both tested settings were shown to produce quality spectra for long pulse beam shots. Short pulse data were also collected; however, the data were not as clean with the exception of the Al data shown. This could be resolved with more filtration to better sort out the hard x-rays produced during high intensity laser shots. The data that were gathered by LLXS during this experimental campaign can be used to determine the temperature and density of the plasma by measuring spectral line ratios and line widths, respectively. The data produced from setting three may also have a wide enough spectral range to determine the plasma temperature from the free-bound continuum as well. This gives a self-consistent means of checking the initial measurement without having to compare the result with a second diagnostic. Each type of crystal used (CsAP, ADP, and PET) produced fine spectra. The data from the ADP crystal are of particular note, as ADP has approximately 5 times lower reflectivity than the other crystals tested for their respective energy ranges.8 The LLXS design can accommodate with a CCD detector sitting ∼30 cm from TCC, therefore it is attached to an airbox with an onboard computer and cooling system. However, after a few days of shots the thermoelectric cooler on the CCD malfunctioned. It is currently hypothesized that the EMP generated from the high flux environment of Orion was too great for the current shielding on the CCD. Additional shielding will be implemented on further campaigns. ACKNOWLEDGMENTS
This work performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract No. DE-AC52-07NA27344. 1 W.
L. Bragg, Proc. Camb. Philos. Soc. 17, 43 (1912).
2 A. H. Compton and S. K. Allison, X-ray in Theory and Experiment (D. Van
Nostrand Company, New York, 1935). U. Akli et al., J. Instrum. 5, P07008 (2010). 4 D. J. Hoarty et al., High Energy Density Phys. 9, 661 (2013). 5 D. J. Hoarty et al., Phys. Rev. Lett. 110, 265003 (2013). 6 N. Hopps et al., “Overview of project Orion,” in Proceedings of the SPIE 7916, High Power Lasers for Fusion Research, 79160C, 18 February, 2011. 7 B. L. Henke, E. M. Gullikson, and J. C. Davis, “X-ray interactions: Photoabsorption, scattering, transmission, and reflection at E = 50–30000 eV, Z = 1–92,” Atomic Data Nucl. Data Tables 54(2), 181–342 (1993). 8 A. J. Burek, Space Sci. Instrum. 2, 53–104 (1976). 3 K.
FIG. 6. A target of 790 nm KCl sandwiched between two layers 310 nm of Parylene-N was shot with two 800 ps long pulse beams. Spectra from both PET crystals were combined to form one continuous spectrum.
Review of Scientific Instruments is copyrighted by the American Institute of Physics (AIP). Redistribution of journal material is subject to the AIP online journal license and/or AIP copyright. For more information, see http://ojps.aip.org/rsio/rsicr.jsp
