Converting existing optical detectors into fast x-ray detectors

The very short burn time and small size of burning plasmas created at advanced laser-fusion facilities mean that diagnostics with high time resolution of a few picoseconds and diagnostics with high spatial resolution will be required for ignited plasma research. These instruments must function in an environment of extremely large neutron fluxes that will cause conventional diagnostics to fail because of radiation damage and induced background levels.¹ A solution to this challenge is to perform an ultrafast conversion of the x-ray signals into the optical regime before the neutrons are able to reach the detector and then to relay image the signal out of the chamber and into a shielded bunker, protected from the effects of these neutrons. The relay imaging could be similar to the velocity interferometry system for any surface reflector (VISAR) diagnostics implemented on the National Ignition Facility, the Z machine at Sandia National Laboratory, and the OMEGA laser. This technique performs an ultrafast conversion of the x-ray signals into the optical regime using the small changes in refractive index produced by the absorption of ionizing radiation in semiconductors. The x-ray image is encoded in at most a few ps on an optical probe beam traveling through a 50–100 μm thick semiconductor. The optical probe beam can be conveniently relay imaged to a remote shielded location for recording. The all-optical conversion process involves no electrical devices, so it is highly resistant to electromagnetic pulse (EMP) effects. Because the signal conversion to optical light is nearly instantaneous, even the effects of neutrons hitting the Hohlraum wall (it takes ∼60 ps for 14.1 MeV neutrons to reach the Hohlraum wall) and target positioner should not affect the measurement and the large effects of neutrons interacting with the chamber and target area are completely avoided due to the differences in time scales for the neutrons to travel several meters. The 1–2 ps time resolution possible with this technique is comparable to or better than the fastest electronic devices, the dilation x-ray imager, DIXI, camera,² and or fast streak cameras, and the 5–10 μm detector spatial resolution possible³ requires far less magnification to achieve a given spatial resolution at the target than the ∼200 μm detector resolution of the fast electronic DIXI or streak cameras. A hardened single line-of-sight camera designed for the NIF can reach the beginning of the burning plasma regime at yields less than 5 × 10¹⁶ with 30 ps and 10 µm temporal and spatial resolution, respectively.⁴ 

Ultrafast picosecond diagnostics based on these principles can be developed to be used for x-ray imaging and to measure x-ray spectra using pulsed or spectrally chirped optical probe beams to encode gated or streaked x-ray images on an optical probe beam.^3,5 Multiple time snapshots, short x-ray movies with 4–20 frames and ps resolution, can be taken from a single line-of-sight by probing the semiconductor with multiple optical beams separated by a combination of probe beam polarization, angle, and wavelength.^3,6 In addition, it may be possible to make measurements driven by the neutrons from the target at late times close to or even after the neutrons hit the chamber wall with this technique, which could not be done in any other way. Multiple gated images and continuous record (streaked) x-ray data with ps time resolution can be recorded without exposing any expensive electronic equipment to the effects of neutron generated EMP and neutron radiation.

Existing inertial confinement fusion facilities already have optical diagnostics that are capable of being converted into high speed x-ray diagnostics. The OMEGA laser facility, for instance, has the OMEGA high resolution velocimeter (OHRV) diagnostic that is capable of high spatial resolution and sampling the phase change in a sample with a few picosecond time resolution between the two probes.⁷ The OMEGA laser facility also has the VISAR diagnostic that provides one-dimensional imaging with continuous time resolution.⁸ The Sandia Z facility also has a one-dimensional spatial imaging VISAR diagnostic and a photonic Doppler Velocimeter (PDV) system.⁹ Each of these diagnostics has the potential to be converted for use as x-ray diagnostics.

Experiments were conducted at the Laboratory for Laser Energetics (LLE) in Rochester, NY, using the OMEGA laser. The emphasis on this project was to make changes to existing diagnostics on the OMEGA laser facility to enable the demonstration of the concepts described above. That is to show that x-ray images could be made by imaging an x-ray source onto a semiconductor to create a spatially dependent index of refraction change in the semiconductor. This spatially dependent index of refraction change could then be detected via an optical instrument such as the VISAR or the OMEGA high resolution velocimeter (OHRV)¹⁰ diagnostics at the OMEGA laser facility.

The x-ray source used to demonstrate the imaging diagnostics was formed by imploding a backlighter capsule.¹¹ The backlighter capsule itself was a 9 µm thick, 860 µm diameter CH capsule with no fill gas. A total of 39 beams with SG4 phase plates were used to implode the capsule, thereby creating the x-ray source. The 39 beams had a combined energy of 19.5 kJ and utilized a laser pulse shape, which was a 1 ns square pulse in time, listed as the pulse shape SG10vA01. The resultant x-ray emission from the backlighter capsule was imaged with an x-ray framing camera (XRFC) and with the VISAR and OHRV diagnostics. The DANTE diagnostic was also used to measure the x-ray power as a function of time.¹² A schematic of the experimental setup with the backlighter capsule and the geometry of the diagnostics is shown in Fig. 1. The angular orientations of the various 10 in. manipulators (TIMs) can be seen in Ref. 13.

The XRFC was used to capture two-dimensional images of the imploding backlighter capsule. The XRFC was located on the TIM3, which allows researchers the flexibility to place a variety of diagnostics inside the vacuum chamber, and had a 16 pinhole array with a magnification of 12. The pinholes were 6 µm in diameter and had 100 µm thick Be filtering with the XRFC 38.1 cm from the backlighter capsule. Figure 2 shows the 2D x-ray images captured by the XRFC. At ∼600 ps, the capsule size is 310 µm in diameter and the capsule size at ∼1.4 ns is ∼150 µm. Peak emission occurs between 1.3 and 1.4 ns as seen in the DANTE diagnostic.

The x-ray emission from the backlighter capsule was imaged onto a semiconductor, diamond or quartz, using a miniature x-ray snout, as shown in Fig. 3. The miniature x-ray snouts were made via additive manufacturing with a tantalum slit array (VISAR) or pinhole array (OHRV) glued to the end of the snout to provide the imaging onto the semiconductor glued to the back of the snout. The slit was located 5 mm from the backlighter capsule such that the emission was magnified a factor of two onto the semiconductor placed 15 mm from the backlighter capsule. Each snout was held by a stalk in the TIM listed in Fig. 1.

The x-ray emission from the backlighter capsule was imaged onto a semiconductor, diamond or quartz, using a miniature x-ray snout. The VISAR then probed the semiconductor from the back and measured the induced phase change inside the semiconductor due to the impinging x rays. The raw effect of the spatial phase change in the semiconductor on the VISAR diagnostic can be seen in Fig. 4. We also saw some blanking of the semiconductor due to too many x rays hitting the semiconductor at peak emission. Phase unfolds of the VISAR as a function of space and time are shown in Fig. 5. The DANTE diagnostic was also used to measure the spatially integrated x-ray emission from the capsule as a function of time. A comparison between the VISAR and DANTE diagnostics is shown in Fig. 6. Specifically, a lineout of the probe’s phase change after passing through the semiconductor measured by the VISAR is compared to the DANTE diagnostic, channel 11.

We obtained x-ray images on the VISAR for both diamond and quartz on all three shots (diamond was used on two of the shots). The quartz interferogram had a smaller phase shift than the diamond semiconductors. The VISAR was operating as a time differential interferometer with the two probes delayed by an etalon such that one probe sampled the semiconductor at a fixed time delay relative to the initial probe beam. As such, the etalon chosen for the VISAR dictates the time response of the measurement. The thinnest etalon used provided a time difference of 10.3 ps. Although the spatial resolution was not measured, the imaging slit used would set the minimum spatial resolution at ∼15 µm. The OHRV diagnostic was mistimed on the first two shots and blanked on the last shot and so unfortunately did not get two-dimensional spatial information. Running both the VISAR and OHRV diagnostics simultaneously led to long shot cycles (∼1.5 h) due to the iterative alignment of the x-ray snouts and the VISAR and OHRV return signals from the semiconductors. The x-ray snouts were made by additive manufacturing, and perhaps because of this, there was some difficulty with tilt angles on aligning the diagnostics. Future work would have to reduce the sensitivity to alignment perhaps by incorporating the semiconductor into a retroreflector configuration and making more precise x-ray snouts with better shielding for the OHRV diagnostic.

The DANTE channel 11 with its solid angle of 7.07 × 10⁻⁶ sr and responsivity of ∼3 × 10⁴ V/GW measures ∼120 J of x rays, within its spectral width of ∼2–6 keV, at the source over the ∼100 ps FWHM near peak emission. With a 10 µm wide by 500 µm long slit located 5 mm away, the fraction of x rays passing through the slit would be [0.01 * 0.5/(2π5²)] = 3.2 × 10⁻⁵. That would pass up to 3.8 mJ of x rays through the slit over ∼100 ps and that spectral range. The image of the FWHM of the x rays would cover 0.15 mm * M by 0.5 mm * (M + 1), which with M = 2 would be 0.45 mm². The fluence on the semiconductor would be 3800 μJ/0.45 mm² or 8400 µJ/mm². Over 10 ps, that energy would be reduced to 840 µJ/mm². The measured peak phase shift of 1.4 rad would then imply a responsivity in the diamond semiconductor of (1.4 rad)/(840 µJ/mm²)/2 or 0.0008 rad mm² μJ⁻¹.

It is desirable to have a faster x-ray diagnostic to measure the burn width and bang time of magnetically driven inertially confined fusion implosions on the Sandia Z machine. Although one solution to that would be to build a streak camera instrument for that purpose like the SPIDER diagnostic¹⁴ built for the National Ignition Facility (NIF), the Z facility offers additional challenges due to the unique chamber, target layout, diagnostic access, and energy. The diagnostic would have to look along the equator, which is below the water line of the tank containing the transmission lines that surround the vacuum chamber. As such, the diagnostic would have to be placed in a “boat” and the large shocks, which occur after the Marx banks are fired and the energy is delivered to the target, would likely destroy the photocathode on each shot. To prevent this from occurring, an elaborate shock absorber structure would have to be engineered.

One alternative to measuring the x-ray burn width is to adapt the existing photon Doppler velocimeter (PDV) diagnostics. This would simply involve using a semiconductor to absorb the x rays from the implosion and to measure the change in the index of refraction of the semiconductor with the existing PDV system. The PDV systems generally use 1500 nm probe wavelengths. The semiconductor therefore has to be chosen to have its bandgap energy at a higher energy than the probe beam energy. The semiconductor should also have a high Z to absorb the x rays. One candidate is InP, which has a bandgap of 1.27 eV (976 nm) and a high x-ray absorption per unit length.

Semiconductors normally have a fast rise time, ∼100 fs, but a slow fall time, several hundred picoseconds to nanoseconds. The carrier lifetime in semiconductors, however, can be reduced in various ways. One way is to create damage centers inside the semiconductors using proton or neutron bombardment. Another way is to add dopant to the semiconductor. The response time of InP with various dopants added has been studied.¹⁵ The authors of Ref. 15 found that a sulfur concentration of 1.9 × 10¹⁸ cm⁻³ added to InP gives a decay time of 6.5 ps. This concentration level of S dopant in the InP was the starting point of the design presented in this article. Higher levels of S dopant, however, will decrease the decay time to sub-picosecond. When coupled to a chirped PDV system, the response time can be one to a few picoseconds, an order-of-magnitude faster than the streak camera-based Spider diagnostic on the NIF.

The x-ray probe is then simply a S-doped InP semiconductor with a thin layer of aluminum, 100 Å, on the side facing the x-ray source to keep visible and ultraviolet light from creating electron–hole pairs in the InP. An additional filter can be used to reduce the x-ray flux onto the semiconductor. These are then glued onto a Thorlabs fiber polishing puck, which holds the APC fiber at the same angle as the semiconductor, as shown in Fig. 7. The other end of the fiber is connected to the PDV system. This probe end is destroyed on each shot but is made up of relatively inexpensive components.

The x-ray end probe was built and implemented on a Z-pinch implosion with their conventional PDV system, not the chirped system. A phase change was detected in the S-doped InP semiconductor. Future tests will be conducted and the results will be compared with gamma-ray reaction history (GRH) and chemical vapor deposition (CVD) diamond detectors. The x-ray end probe will eventually be implemented with the chirped PDV system to enable its fast, sub 10 ps, response time.

The OMEGA laser facility was used to demonstrate high-temporal-resolution x-ray imaging by using an x-ray snout to image an imploding backlighter capsule onto a diamond or quartz semiconductor. The semiconductor was simultaneously probed with the existing VISAR diagnostic, which uses an optical streak camera and provided a one-dimensional image of the phase in the semiconductor as a function of time. This approach would allow a sacrificial semiconductor to be attached at the end of an optical train with the VISAR and optical streak camera placed in a shielded bunker to operate in a high neutron environment and obtain time-dependent one-dimensional x-ray images or time-dependent x-ray spectra from a burning plasma.

It is desirable to have a faster x-ray diagnostic to measure the burn width and bang time of magnetically driven inertially confined fusion implosions on the Sandia Z machine. As such, an x-ray probe was designed to interface with the existing conventional PDV system at Sandia. The x-ray end probe was built and implemented on a Z-pinch implosion. A phase change was detected in the S-doped InP semiconductor. Future tests will be conducted and the results will be compared with gamma-ray reaction history (GRH) and chemical vapor deposition (CVD) diamond detectors. The x-ray end probe will eventually be implemented with the chirped PDV system to enable its fast, sub 10 ps, response time.

We wish to thank the OMEGA and Sandia operations team. This paper was prepared by LLNL under Contract No. DEAC52-07NA27344. This document was prepared as an account of the work sponsored by an agency of the United States government. Neither the United States government nor Lawrence Livermore National Security, LLC, nor any of their employees, makes any warranty, expressed or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States government or Lawrence Livermore National Security, LLC. The views and opinions of the authors expressed herein do not necessarily state or reflect those of the United States government or Lawrence Livermore National Security, LLC, and shall not be used for advertising or product endorsement purposes.

The data that support the findings of this study are available from the corresponding author upon reasonable request.