Gamma-ray measurements for inertial confinement fusion applications

Understanding the nuclear phase of an inertial confinement fusion (ICF) implosion is necessary to improve ICF capsule design^1–4 and study the recently achieved igniting plasma condition.^5–9 Nuclear diagnostics have played a critical role in even the early stage of ICF development and are getting more important as nuclear yields are increasing. Examples of nuclear diagnostics include nuclear activation detectors,^10–12 neutron time-of-flight spectrometers,^13–15 recoil particle spectrometers,^16–20 neutron imaging systems,^21–23 and radiochemical diagnostics.^24–27 

Gamma rays generated from ICF plasmas provide unique and important information for studying the dynamics of the nuclear phase of the ICF process because the energy spectra and amount of γ rays produced are a direct signature of the nuclear fusion reactions themselves. For example, the 16.75 MeV γ-ray line originating from the nuclear fusion reaction between deuterium (D) and tritium (T) can offer reaction history information, such as the time of peak reaction rate (i.e., bang time) and how long thermonuclear conditions are maintained (i.e., burn width).^28–31 Due to their high-energy and penetrating nature, γ rays are the most unperturbed fusion products, which can preserve the original birth information of the fusion process even from the highly compressed fuel conditions achievable at the National Ignition Facility (NIF). In addition, experimental observables through γ rays are essentially averaged over all directions about the ICF implosion, providing a global reference for the line-of-sight-specific measurements typical of x rays and neutron diagnostics.

Since the early period of ICF research on the NOVA Laser Facility, fusion γ-ray measurements were conceived of as a direct measure of nuclear reaction rates (unlike x rays) without being compromised by Doppler spreading (unlike neutrons).^32–34 However, unambiguous γ-ray measurements for ICF study further required a decade-long period of technological development, which included a deepening understanding of fusion γ-ray characteristics and innovations in instrument performance. Recently, a γ-ray-based reaction history instrument has achieved a temporal response of 10 ps and, therefore, was able to accurately measure the first sub-100 ps burn duration (i.e., 88 ± 15 ps), which occurred during the first ignited cryogenic-layer shot at NIF (i.e., N210808).^7–9 Now, γ-ray measurements of burning plasmas can bolster the continued improvement in simulation codes, so that the impediments to ignition can be truly identified and addressed.

In this article, the production mechanisms of the prompt and secondary γ rays are presented in Sec. II. Various ICF performance parameters that can be derived from γ-ray measurement are introduced in Sec. III. Sections IV and V describe the γ-ray detectors fielded or proposed, especially for high-yield ICF experiments at the Omega Laser Facility and NIF. Over the past few years, γ-ray technologies developed primarily for ICF research have been extended beyond the initial ignition campaign. Section VI introduces a few examples of non-ICF applications of Cherenkov-based γ-ray detectors. Finally, challenges and future directions of γ-ray measurements are discussed in Sec. VII.

A mixture of deuterium and tritium is mostly used in ICF experiments because the DT fusion reaction is achievable at the lowest plasma temperatures. Meanwhile, a variety of other fuel species are often utilized to study specific aspects of the implosion process. For example, the use of ³He as a surrogate for tritium provides an alternative way of characterizing capsule performance while minimizing complications associated with tritium filling or handling. Table I lists examples of nuclear fusion reactions widely used in ICF research, especially those producing high-energy γ rays.

Although primary γ rays generated from fusing fuels provide a direct indication of fusion production, the fusion products such as 14.1 MeV neutrons from D + T fusion reactions can generate additional γ rays via inelastic scattering on nuclei. For example, the intensity of the 4.44-MeV line from inelastic scatter off carbon in a plastic ablator or diamond-like carbon ablator [i.e., ¹²C(n,n’γ)] is proportional to the product of the capsule’s fusion yield and the ablator’s areal density.⁴⁷ Therefore, the neutron-averaged areal density (ρR) of the remaining ablator near bang time can be inferred from the 4.44 MeV γ-ray yields. These ablator γ rays are considered “prompt” since the neutron transit time across the compressed capsule is only a few ps (whereas fusion burn widths and diagnostic temporal resolution are typically on the order of 100 ps).

Fusion γ rays or neutron-induced ablator γ ray signals can be contaminated by the inevitable neutron-induced background γ rays generated when the 14.1 MeV fusion neutrons interact with the surrounding materials such as (1) hohlraum (e.g., Au or U) and (2) thermomechanical package (e.g., Si or Al) at NIF.^48,49 Figure 1 shows the calculated γ-ray spectrum including primary and secondary reactions expected during indirectly driven, cryogenically layered DT implosions on NIF. The Monte-Carlo N-Particle (MCNP⁵⁰) simulation used an instantaneous source of 14.1 MeV neutrons reacting with a 0.9 scale NIF hohlraum (93.5 mg Au) and the surrounding thermomechanical package (TMP, 144.8 mg Al) including silicon rings (66.9 mg Si).^48,51 Temporal separation between the prompt fusion and ablator γ rays and the secondary background γ rays from the surrounding materials is in the order of 100 ps due to the neutron transit time across the geometry of the hohlraum and thermomechanical package. Therefore, a fast γ-ray detector helps to isolate fusion γ rays from the background γ rays.

Since the NOVA and Omega Laser Facilities began ICF experiments, thermonuclear reaction rates as a function of time, or burn histories, have been measured in order to constrain implosion simulation parameters such as shell velocity and confinement time. Figure 2 shows schematics of temporal profiles of laser power and resultant reaction rate, where bang time is defined as the time from the beginning of the laser rise (e.g., 50% rise in the foot of the pulse) to the peak of reaction rate. Bang time is important for benchmarking simulations because it is a measure of energy absorbed by the capsule, and oftentimes, simulations are forced to match the observed bang time by adjusting laser drive factors.^52,53

Traditionally, burn history has been measured by using a plastic scintillator placed close to the target chamber center (TCC) and relaying light generated by inelastically scattered 14.1 MeV fusion neutrons. The stand-off distance of the scintillator needs to be on the order of few cm to minimize time-of-flight spreading due to the thermal Doppler effect. On the NIF, however, diagnostics need to remain at much larger distances from TCC (>30 cm), which makes it hard to achieve a high-bandwidth reaction history measurement relying on fusion neutrons. For a 4 keV DT plasma ion temperature, for example, the time-of-flight spreading will exceed 12 ps for a neutron detector placed greater than 5 cm from TCC, increasing proportionally with detector distance and the square root of ion temperature.^14,54 This time-of-flight spreading can begin to interfere with burn width and features in the reaction history but becomes quite useful at large distances for neutron-time-of-flight (nToF) measurement of ion temperature.

Fusion γ rays such as DTγ in DT implosions and HTγ in THD (tritium-hydrogen-deuterium) implosions are an alternative approach for measuring reaction history since the γ rays travel at the speed-of-light and are not subjected to time-of-flight spread.^55–57 The Gamma Reaction History (GRH) diagnostic—to be explained in Sec. IV in detail—was developed for NIF to convert MeV γ rays into UV/visible photons for high-bandwidth optical detection.

Although fusion neutrons are typically considered the gold standard as a yield signature for any neutronic fusion reaction, highly compressed ICF conditions cause complications in accurately measuring the total fusion neutron yield. The compressed capsules are so dense that the 14.1 MeV DT fusion neutrons scatter on their way out of the capsule and no longer represent a reliable mono-energetic peak for total yield fusion diagnostics. For example, an MCNP calculation for a compressed cryogenically layered ICF capsule (i.e., 800 mg/cm² of DT fuel and 400 mg/cm² of plastic ablator areal density) estimates that about 20% of 14.1 MeV neutrons are down scattered.⁵⁸ Furthermore, because of asymmetries in the compressed capsule, the amount of neutron scattering differs depending on the line of sight. Many neutron detectors at various locations on the NIF are used to measure both the un-scattered and the scattered neutrons and then are averaged to get a total yield measurement.

In contrast, the 16.75 MeV DT fusion γ rays are not nearly as attenuated (∼3%) as 14.1 MeV fusion neutrons (∼20%) based on the MCNP simulation and, therefore, can provide a reliable alternative for total fusion yield measurements.⁵⁸ Gamma rays offer the potential of having one detector in one location giving the spatially averaged total yield, avoiding the complications of the neutron measurements. Meaney et al.⁵⁸ demonstrate that the 10 MeV threshold γ-ray-based yield measurement is strongly correlated with independent neutron yield measurements on NIF. Directly driven exploding pushers, having negligible areal density, are used for in situ cross-calibration of a γ-ray detector against existing neutron yield measurements. In this way, γ rays can be used to infer total DT neutron measurements on indirectly driven, cryogenically layered implosions. In addition, using γ rays in combination with neutrons for yield measurements offers a potential for improved precision of physics values, such as the total down scattering fraction (TDSF), which is a direct measure of total areal density (i.e., fuel plus ablator). The TDSF = 1 – [Yield (13–15 MeV neutrons)/Yield (16.75 MeV γ rays)], where Yield (13–15 MeV neutrons) is the unscattered portion of the neutron yield measurement, and Yield (16.75 MeV γ rays) is the total fusion neutron yield inferred from the γ ray measurement.

In ICF, the areal density (ρR) of an imploding spherical ablator is an important design parameter to achieve hotspot ignition. Reaching high fuel areal density depends on precise control of the symmetry and stability of the imploding ablator shell. Although the areal density of the fuel has been measured by various techniques such as down-scattering of neutrons^17,18 and knock-on production of charged fusion particles,^59,60 there are only a few techniques in measuring the areal density of an ablating shell. On NIF, two-dimensional x-ray radiographic techniques provide the in-flight shape of imploding capsules.⁶¹ Through the sequence of x-ray images, implosion velocity and density asymmetries can be obtained.

In a complimentary way, by measuring the 4.44 MeV characteristic γ-ray line [i.e., ¹²C(n,n’γ)], the GRH team has successfully completed a precise analysis of the ablating shell areal density and also inferred the movement of the remaining ablator shell relative to the time of peak fusion rate.^62–67 It is observed that the ablator shell is often still moving inward at fusion bang-time as evidenced by an increasing carbon γ-ray signal at bang time, indicating that the kinetic energy of the ablating shell does not couple to the fuel efficiently.⁶⁸ The relative timing between fusion bang time and ablator peak time provides useful diagnostic information to help optimize energy coupling from the remaining shell to the fuel. It is expected that ablator peak time will proceed the fusion bang time in an ignited capsule as the alpha burn wave continues to propagate while the capsule is disassembling.

MeV γ rays present a challenge for detection and characterization. Typically, high-Z inorganic scintillating crystals (e.g., NaI, CsI, BGO, …) are used in counting mode to detect γ rays in nuclear physics experiments.⁶⁹ Energy deposited in the crystal by a single γ-ray interaction is characterized by the amount of light generated in the crystal and correlated to the incoming γ-ray energy. However, the flux of γ rays generated by a short-burst ICF implosion easily overwhelms crystal detectors unless they are placed hundreds of meters away and arrayed in large numbers so as to provide adequate statistics. Instead, for ICF experiments, it is preferred to use the gas Cherenkov mechanism in the current mode. Cherenkov radiation is produced whenever charged particles travel faster than the local speed of light in a medium (e.g., gases, liquids, solids).⁷⁰ It is the blue glow typically associated with anything nuclear. The most obvious example is the glow within a water cooling pool about a small nuclear reactor or spent nuclear fuel. The fission products contained within often undergo beta decay in which the emitted electron is relativistic and travels faster than the speed of light in water [v_c = c/n, where c is the speed of light in vacuum and n is the index of refraction of the water (n_water ≈ 1.3)]. Cherenkov radiation can be thought of as the optical analog to a supersonic acoustic wave (i.e., a shock wave), in which an object traveling faster than the speed of sound causes the sound waves to pile up and create a sonic boom. Gas Cherenkov detectors (GCD) take advantage of this mechanism by converting γ rays to relativistic electrons in a convertor plate in the front of a pressurized gas cell. The material of the convertor plate is typically chosen to be low-Z (e.g., Be, Al, or C in the form of graphite) such that Compton scattering is the preferred conversion process, preserving most of the energy and direction of the incoming γ rays in conversion to Compton electrons (as opposed to electron production through the photoelectric effect or pair production).⁷¹ The index of refraction of the gas is determined by the selection and the density (or pressure at a fixed temperature) of the gas.

Figure 3 describes the conversion processes that occur inside a GCD. DT fusion γ rays from an ICF implosion impact the γ-to-electron converter in the nose cone of the detector. The majority of γ rays pass through undisturbed, but a small fraction (∼10⁻³) convert via Compton scattering to a single electron of comparable energy and direction, as previously mentioned.^72–74 The resulting relativistic electrons travel through the pressurized gas within the cell, producing multiple Cherenkov photons per electron in the ultraviolet and visible parts of the spectrum. This optical radiation is detected on the back end by a photomultiplier tube (PMT), but this PMT must be shielded from direct exposure to x-ray radiation also coming from the implosion. The UV/visible light must also be efficiently collected; hence, a Cassegrain telescope configuration is used to allow direct line-of-sight shielding in the form of a tungsten block and collection optics in the form of primary and secondary mirrors. All this fits within a tight cylindrical package, which can be conveniently inserted into an ICF target chamber to get close to the implosion to maximize solid angle collection (typically 20 cm at Omega). The UV/visible photons are converted to electrons at the photocathode at the front of the PMT, amplified by the microchannel plate (MCP) within, and collected as a transient current at the anode at the back end of the PMT. This PMT signal is then sent over a coax cable to a fast digitizer for recording the current as a voltage across a 50 Ω load as a function of time. In all, the conversions proceed from γ rays, to relativistic electrons, to UV/visible Cherenkov photons, to photoelectrons (photocathode), and to secondary electrons (MCP), which are collected at the PMT anode. Additional conversions are often employed in the form of electrical-to-optical, and vice versa, using Mach–Zehnder interferometers and fiber optics in order to preserve the bandwidth that might, otherwise, be lost over long coaxial signal cable runs.^75,76

Figure 4 shows a calculated temporal response for the gas cell assuming an impulse of γ rays (i.e., a delta function source of γ rays in time being launched into 4π from a point source located 20 cm from the detector nose). The Cherenkov mechanism is inherently fast, with just over 10 ps of temporal dispersion for the resulting bunch of Cherenkov photons reaching the photocathode of the PMT. This is much faster than even state-of-the-art PMTs containing MCPs (as opposed to dynodes), which typically add at least another 100 ps of dispersion.⁷⁷ For this reason, the PMT is typically the bottleneck in maintaining high bandwidth. This was the primary driver for developing the pulse dilation PMT that will be discussed in Sec. IV C.

Figure 5 depicts the calculated Cherenkov threshold energy for incoming γ rays, below which no Compton electrons are generated with high enough energy to produce Cherenkov photons.⁷⁸ A favorite choice for detecting fusion γ rays is carbon dioxide (CO₂) because it is safe and convenient to work with and provides a range of n at reasonable pressures (e.g., <400 psia) appropriate for detecting 16.75 MeV fusion gammas. At 100 psia of room temperature CO₂, the threshold energy below which γ rays will not produce Compton electrons with sufficient energy to generate Cherenkov light is about 6.3 MeV. To drop the energy threshold low enough to detect the 4.44 MeV carbon γ rays resulting from the inelastic scatter of 14.1 MeV neutrons off plastic or diamond ablators, it is necessary to switch to a fluorinated gas, such as SF₆ or C₂F₆ (as shown in Fig. 5).

Several iterations of the GCD have been developed for use at the Omega Laser Facility over the past 2 decades. GCD-1 is made of an aluminum body with o-ring seals designed for up to 100 psig of gas pressure and is only approved for non-fluorinated gases (e.g., CO₂) due to the potential concern for fluorine leaks into the target chamber poisoning the tritium recovery beds. GCD-2 was a redesign to couple the optical signal out the back of the gas cell using relay optics to a streak camera nearly 10 m away in an attempt to eliminate the PMT bandwidth bottleneck. However, due to poor optical collection efficiency, the wasted signal in coupling a round photon bunch to a narrow streak slit, and the high background sensitivity of the streak camera phosphor and a charge-coupled device (CCD), a definitive high-bandwidth fusion γ-ray signal was never conclusively demonstrated using GCD-2. GCD-3 was a redesign of GCD-1 to allow higher pressure operation of up to 400 psig and the use of fluorinated gases.^79–82 This was accomplished by using explosively welded aluminum-to-stainless steel flanges, which allowed for metal gasket on stainless knife edge seals to essentially eliminate the potential for leaks, while also keeping the gas cell body lightweight aluminum. The initial driver for GCD-3 capabilities was the desire to conduct nuclear astrophysics studies on an ICF platform by measuring the 5.5 MeV γ rays resulting from H + D fusion (an important step in the solar pp fusion chain).⁴² GCD-1 was only capable of reducing the threshold down to 6.3 MeV, so was blind to the HD γ rays. GCD-3 opened a whole new part of the γ-ray spectrum down to nearly 2 MeV. GCD-3 has since been duplicated for dedicated use on NIF where it sits inside a reentrant diagnostic insertor allowing it to get within 4 m of TCC. Figure 6 shows GCD-1 (foreground), and its successor, GCD-3 (aft), laying on the upper deck at Omega, as they are being swapped out of the Ten Inch Manipulator (TIM), which can be seen as the rectangular box with open lid pointed at a downward angle toward TCC.

In addition to providing bang time and burn width, GCD-1 on OMEGA was actively used to study characteristics of fusion γ rays. Although charged particle branches of nuclear fusion reactions have been extensively studied in the past, additional γ-ray branches are not well known due to their exceptionally small cross-sections. D-T γ rays shown in the first row of Table I are no exception. An R-matrix nuclear analysis⁸³ has predicted two branches of D-T γ rays [i.e., the higher energy γ rays (γ₀) due to the transition to the ground state, and the lower-energy γ rays (γ₁) due to the transition to the intermediate first excited state of ⁵He]; however, their relative intensities (i.e., γ₁/γ₀) have not been calculated. Determining the spectral shapes of D-T fusion γ-ray emissions has long been a subject of study using beam target accelerator-based experiments; however, there have been large inconsistencies, especially when determining the intermediate excited state due to the large neutron-induced background issue. Using a GCD-1 on OMEGA, the team from the Los Alamos National Laboratory (LANL) and the Atomic Weapons Establishment was able to confirm for the first time the presence of two γ-ray branches and reported a ratio of γ₁:γ₀ = (2.1 ± 0.4): 1 and eliminate ambiguity in the fusion γ-ray measurement.^84, Figure 7 shows the resultant D-T γ-ray energy spectrum. Historically, the DT γ-to-neutron branching ratio was reported with a large uncertainty (x30), and only measured at particle accelerator facilties.^{37,38,120,121,124,125} Under ICF conditions, the total D-T γ-to-neutron branching ratio [D(t, γ₀+γ₁)⁵He/D(t, n)⁴He] was determined to be (4.2 ± 2.0)×10^-5 using a GCD-1,^85,86 (4.6 ± 0.6) ×10^-5 using a GCD-3,¹²³ and (8.42 ± 2.84)×10^-5 using a Diagnostic for Areal Density (DAD).³⁶ ICF measurements were done at ion temperatures of (5 − 19) keV, those are quite low compared to previous accelerator measurements.^36,85,86,123 In a practical sense, precise measurements of the branching ratio D(t, γ₀+γ₁)⁵He relative to D(t,n) ⁵He are important in order to diagnose the resultant fusion yield of cryogenically layered implosions at the NIF.

As consideration was being given to Reaction History measurements on NIF prior to the 2010 start of the National Ignition Campaign (NIC), a primary concern was the shortage of diagnostic insertors and manipulators (DIM⁸⁷). NIF was to start operations with only two DIMs. Multiple neutron, charged particle, and x-ray diagnostics needed to be close to TCC. The much larger laser energies and debris-forming hohlraum/thermomechanical package used for indirect drive on the NIF meant that the standoff distance for diagnostics needed to be larger than what is accessible at Omega; typically ≥30 cm. This precluded the use of a Neutron Temporal Diagnostic (NTD)⁸⁸ for Reaction History due to the unacceptably large Doppler spreading of the neutron signal at this distance, and a GCD would have taken up the majority of the payload space and weight limitations of a single DIM. Instead, the effort was refocused toward developing a gas Cherenkov detector that would be adequately sensitive operating outside of the 5 m radius target chamber. Being outside the chamber significantly relaxed the volume and weight constraints on the diagnostic and allowed the design of more efficient collection optics utilizing off-axis parabolic mirrors (OAPs). It also enabled the mounting of multiple cells, which could be run at various energy thresholds (i.e., gas pressures) in order to conduct crude spectroscopy. The primary intention of this spectroscopic capability was the ability to separate fusion, carbon, and hohlraum/TMP γ rays from each other, but had the added benefit of enabling other nuclear plasma physics spectroscopic studies in the process.

Figure 8 shows the Gamma Reaction History (GRH) diagnostic as conceived in 2008. The gas cell is mounted outside on a target chamber port at 6 m from TCC and the Cherenkov light generated within is redirected and focused down onto the PMT in a folded path. The desired Cherenkov light passes through the sapphire pressure window after being turned 180° such that any unwanted Cherenkov produced in the pressure window goes in the opposite direction.^89–92 The PMT, also mounted in the opposite direction, is easily shielded with a large block of tungsten alloy (W). Fiber optic light insertion through the flat turning mirror allows for the injection of timing fiducial and calibration signals. The much longer signal transmission distance at NIF as compared to Omega (i.e., ∼50 vs ∼10 m) necessitated the use of Mach–Zehnder conversion to an optical signal for transmission over fiber optic and optical-to-electrical conversion at the digitizer in the diagnostics mezzanine.^75,76

The GRH was first prototyped and installed on Omega for testing in 2009. The following year it was brought to the High-Intensity γ-Ray Source (HIγS), Triangle Universities Nuclear Laboratories (TUNL) on the Duke University Campus, NC, for thorough calibration using a high repetition rate (∼5 MHz), single-hit, energy-selectable, γ-ray signals.^78,93 Figure 9 shows the four gas cells that were installed on NIF and acquired data on the first NIC shot in September 2010. The four square boxes with flexible steel-braided conduit near the centerline of the diagnostic house the Mach–Zehnder for the PMTs, which are well shielded by a stack of eight 1 in. thick tungsten alloy plates mounted within the line-of-sight from TCC. The gold-colored boxes cover the fiber optic insertion ports into each turning mirror. The black cylinders near the blue gunite shield wall on the target chamber are the actuators for gate valves that are left open during shots to allow the gammas to pass through minimal debris shields and vacuum boundary.

Figure 10 depicts some of the earliest GRH data from the NIC campaign (N101030). The comparison of pressurized (red) and evacuated (green) gas cells demonstrated that the optical signal was unambiguously being generated in the gas and was consistent with expectation for the Cherenkov signal from prompt DT γ rays and delayed, inelastic-scatter γ rays coming from diagnostic snouts within the target chamber.

The GRH has taken data on every yield-producing shot since the inception of DT operations on NIF, measuring bang time (to within ∼30 ps) and burn width (to within ∼15 ps). The burn widths for the first decade of operation always exceeded ∼130 ps, whereas the simulations and x-ray measurements were typically shorter (∼100 ps). This led us to question the accuracy of the γ-ray measurements, although calibrations and OMEGA data clearly indicated that GRH was capable of measuring down to ∼100 ps burn widths with the existing PMTs. The temporal resolution of GRH has been limited to ∼100 ps due to the temporal resolution of the current state-of-the-art PMT itself. This drove the development of the Pulse Dilation PMT (PD-PMT), which is in essence a nondimensional, hardened streak camera, allowing temporal response matching that of the gas cell (∼10 ps).⁹⁴ Over the course of fielding the PD-PMT on NIF (Fig. 11), two sources of background signals were observed (i.e., direct interaction of γ rays with the pulse-dilation tube itself, and scintillation from the gaseous Cherenkov medium).^95–97 Overcoming such background issues, burn width measurements from GCD-3 coupled to the PD-PMT on NIF started in 2017 and remained consistent with GRH. It was not until the “threshold of ignition” shot N210808 that the yield jumped significantly and the burn width came down to ∼90 ps, the first ever below 130 ps, demonstrating that GRH did, indeed, have the temporal response to measure down to 100 ps as designed and providing a clear indication of a propagating alpha burn wave into the cold dense DT fuel layer.

Since only one Compton electron of comparable energy can be generated per “productive” γ-ray, and using the insight from the Monte Carlo simulation that each Compton electron produces on average 30 Cherenkov photons that reach the photocathode, we can estimate for this case that detection of 100 productive γ rays (requires a neutron yield of $Y n D T = 3 , 000 p h 3.7 × 10 − 8 =8.1×1 0 10 ≈1 0 11 D T N$⁠). We use a more rigorously computed version of this “100 productive γ rays” as the minimum yield required to obtain reasonable statistics for bang time and burn width measurements (as seen in the last row of Table II). Table II summarizes the specifications of time-resolved Cherenkov detectors introduced in Secs. IV A–IV C.

In addition to the continued demonstration of DT-fusion ignition at NIF, a number of DOE’s grand challenges can be explored in the unique high energy density plasma (HEDP) environment at NIF. As examples, experiments at NIF will allow for advanced studies on alpha-particle physics in burning plasmas,⁹⁹ demonstration of aneutronic fusion fuel scenarios,^100–102 stellar nucleosynthesis,¹⁰³ and so on. Of particular importance to the ignition campaign at NIF is alpha-particle transport in DT, and D-³He plasmas as alpha confinement is a critical parameter for achieving thermonuclear burn. In addition, the quantitative spectrometry of high energy (multi-MeV) γ rays released in aneutronic fusion reactions at NIF offers a unique opportunity to determine the achievable reaction rates of future advanced-fuel fusion reactors as well as to study astrophysically relevant nuclear reactions involving capture reactions on excited nuclei. For example, the HEDP drives nuclear reactions responsible for the generation of energy and also the formation of heavy elements while accessing lower-energy regimes than possible with beam-target accelerator experiments. Moreover, in contrast to the nuclei that are neutral in accelerator targets, nuclei in the HEDP are highly ionized like those in the cores of stars. In the neutral target, bound electrons shield the positive charges of interacting nuclei and thus reduce the Coulomb repulsion as compared to ionized stellar plasmas. This requires a correction to beam-target data by estimating the screening effect of the bound electrons. Altogether, HEDP facilities make nuclear astrophysics studies possible on earth.^104–107 A γ-ray spectrometer with high sensitivity and energy resolution has the potential to probe such ICF/HEDP nuclear reactions.

The Gamma-to-Electron Magnetic Spectrometer (GEMS) diagnostic has been conceptually designed to measure the prompt γ-ray energy spectrum during high-yield DT implosions at the NIF.^51,108 The operating principles are (1) γ rays from the imploding target encounter a γ to e⁻ converter; (2) the electrons enter a magnetic field, providing spatial dispersion by momentum selection; and (3) the dispersed electrons enter Cherenkov radiators, where their binned energy is converted to UV/visible photons for detection by photomultipliers. The GEMS is intended to measure the ICF generated γ-ray spectrum from 2 to 25 MeV in 21 energy bins with an energy resolution of 3%–5%. The GEMS’s temporal resolution would be better than 1.5 ns; the dynamic range ≥100; and the signal-to-noise ratio ≥5. Using an γ-to-electron conversion efficiency of 5 × 10⁻⁴ and a magnet efficacy of 80%, the minimum required yields to obtain 100 detectable electrons in the energy bins of interest are ≥5 × 10¹⁴ DT neutrons to measure the 4.44 MeV ¹²C γ rays; ≥2 × 10¹⁵ DT neutrons to measure the 16.75 MeV DT γ -ray line. Such a γ-ray spectrometer shown in Fig. 13 can be utilized to support not only the ignition campaign but also basic nuclear physics on NIF and other fusion facilities.

Multi-MeV photon beams are highly penetrating and thus offer various applications such as non-intrusive, active interrogation of special nuclear materials. Short, but intense pulse capability is desirable for fast radiography and active interrogation applications because it can improve the signal-to-noise ratio. Several research groups have investigated the laser production of high-energy photon beams^109,110 or pulse-power-driven photon beams;^111–113 however, it is still challenging to optimize MeV photon beams because specific applications may require new diagnostics development. For example, the GRH detector concept has been transformed to Aerogel Cherenkov Detector for Cygnus (ACD/C) in order to measure the energy spectrum of x rays that are produced by a rod-pinch plasma device and determine how rod materials and geometry affect the x-ray emission spectrum. The development of the ACD/C detector is part of a broader effort by the DOE/NNSA Science Program to improve the overall understanding of the physics of radiographic x-ray sources, produced by machines such as Cygnus at U1a facility at the Nevada National Security Site. ACD/C shown in Fig. 14 is an energy threshold detector, sensitive to x rays with energy only above a certain threshold determined by the index of refraction of the detection medium (e.g., aerogel or quartz). The energy threshold of the ACD/C was tested in the range of 0.3–6 MeV by employing fused silica (2.2 g/cc) and different densities of silica aerogels (10-430 mg/cc).^114–119 Recently, ACD/C was fielded on Sandia’s Z-facility and diagnosed x-ray background produced by pulse-power-driven ICF experiments.

There are few signatures of the transition to a burning (igniting) plasma. The rapid decrease in the burn duration of the thermonuclear DT fuel in an ICF implosion is perhaps the most unambiguous. The most accurate way to diagnose this is through measurement of the γ rays produced in a rare branch of the D + T reaction. The LANL ICF gamma team has developed nuclear science and technology to measure these fast burn times. Nuclear physics is advanced by the measurement of the DT gamma-to-neutron branching ratio and by confirmation of the theoretically predicted double-peaked DT gamma spectrum, never before measured. The team developed the technology to measure burn width as short as 10 ps by careful design of the gas Cherenkov cell and development of the pulse dilation PMT.

Recent inertially confined fusion experiments at the NIF have created burning plasma conditions for the first time. The record-setting experiments demonstrated a tenfold increase in neutron yield, a doubled ion temperature, a burn duration reduced by nearly 40%, and the largest hotspot burning volume ever observed in the laboratory—all of which are consistent with a thermonuclear burning system. Understanding the basic physics mechanisms of energy deposition from the fusion products (boot-strapping) to the compressed fuel is crucial for improving ignition performance. One example of direct evidence of the boot-strapping effects of alpha particles is the abrupt temperature increase at the deposition location, an increase in the burn rate, and a narrowing of the burn width.¹²⁶ The Gamma Reaction History has been used successfully for non-igniting plasmas for a decade and is ready to be leveraged for future studies of igniting plasmas.

The authors would like to thank everyone who contributed to ICF gamma ray research and provided data to this review article; especially, Steve Batha, Steve Caldwell, Charlie Cerjan, Jenny Church, Jorge Carrera, Scott Evans, Cathleen Fry, Hermann Geppert-Kleinrath, Steve Gales, Colin Horsfield, Gerry Hale, Nels Hoffman, Matthias Hochanadel, Justin Jeet, Justin Jorgenson, Morris Kaufman, Jamie Langenbrunner, Alex Leatherland, Frank Lopez, Kevin Meaney, Joe Mack, Aaron McEvoy, Kirk Miller, Eddie Mariscal, Bob Malone, Tom Murphy, Zaarah Mohamed, Tana Morrow, Mike Rubery, Wolfgang Stoeffl, Tom Sedillo, Mark Schmitt, Michael Springstead, Woody Van Dervort, Lucy Wilson, Carl Young, Kevin Yates, and Alex Zylstra. This work was performed by the Los Alamos National Laboratory, operated by Triad National Security, LLC, for the National Nuclear Security Administration (NNSA) of U.S. Department of Energy (DOE) under Contract No. 89233218CNA000001. This study was supported by the NNSA Office of Experimental Sciences Inertial Confinement Fusion (Campaign-10) Programs.

The authors have no conflicts to disclose.

Yongho Kim: Writing – original draft (lead). Hans W. Herrmann: Writing – original draft (supporting).

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