Dynamic high energy density plasma environments at the National Ignition Facility for nuclear science research

The NIF is presently the largest operating high-energy laser system for research of laser based ICF and HED science. The NIF laser facility and the basic concept of laser driven ICF is described in [15–22], and references quoted therein; a review of the National Ignition Champaign (NIC) 2009–2012 is given in [23]. Through the ICF research a key platform for nuclear science research in HED environments has been established.

1.2.1.1. The laser [15–18]

1.2.1.2. ICF research [23]

NIF experimental capabilities are available to a broader scientific community via calls for proposals for NIF experiments under the Discovery Science Program [25]. NIF experiments are typically executed by already established or new experimental 'platforms' for ICF and HED research. Most of these platforms have been developed as a result of HED and ICF research in areas including main-line programmatic materials properties, radiation transport, and ICF. A NIF experimental platform typically consists of an integrated laser and target configuration, and diagnostic suite capable of providing well-characterized pressure, temperature, implosion, or other environments.

A variety of experiments to study matter at the extremes, including studies of material properties, hydrodynamics, and the interaction of intense radiation fields with matter have been conducted as well and provided new platforms.

Particular samples are then varied and studied with a characterized drive and diagnostic suite. Table 2 shows 8 primary platform categories. Each platform represents a range of different experimental configurations that are grouped based on general characteristics of the selected configuration. Each category is linked to information on the current platform designs and experimental results. To the left are additional links to historic specific platform information that is undergoing revisions [25].

Fundamental science at NIF should leverage these platforms whenever possible to minimize the level of effort to acquire physics data. While novel configurations may be used, new configurations may require significant development and evaluation.

An example where an existing platform that may be employed for nuclear physics measurements is the indirect-drive ignition platform that provides a repeatable high-brightness 14 MeV DT fusion neutron source. This enables high-fluence 14 MeV neutron exposures of activation targets of interest to LLNL programs. The neutron energy flux spectrum is characteristic for this source; it is possible to moderate the neutrons and utilize the lower energy neutrons for target exposure. The neutron source (14 MeV, 100 ps) is used to measure neutron induced cross sections relevant to LLNL programs as well as cross sections for astro-physics utilizing lower energy neutrons (e.g. for s-process studies).

The following provides a short overview of the complexity of the ignition target manufacturing for the NIF. An ignition target consists of many individual precision-manufactured components, as specified by a particular target physics campaign. The targets are both complex and precise due to the requirements for their mission; they are truly engineering marvels in tiny packages. Their production requires an interplay among target designers, materials scientists, and precision engineers. The targets must be designed, fabricated, and assembled with extreme precision, so that they will perform as required by the experiment (usually at high temperature and pressures). An overview of different target types is given table 3. Components must be machined to within an accuracy of 1 μm and material structures and features require a smoothness tolerance approaching 1 nm. The most complicated targets are the indirect-drive cryogenic ignition targets. At the center of these targets is a ∼2 mm diameter capsule that is filled with hydrogen fuel—in most cases, a solid (ice) layer of a 50:50 equimolar mixture of deuterium and tritium in equilibrium with its vapor at 1.3–1.5 K below the triple point at shot time. The capsule and the fuel are concentric and centered inside a high-Z hohlraum. An example of an ignition target is displayed in figure 3.

Zoom In Zoom Out Reset image size

Precision engineering design and applied materials research has enabled the target manufacturing team to meet the stringent ICF requirements for capsule surface finish and positioning, performance at cryogenic temperatures, and robustness during assembly and fielding. Assembly stations, tooling and procedures are available to meet target-positioning requirements, and component design changes are implemented to improve assembly yield and throughput [36]. The target technology research and development is continuous by necessity to maintain the ability to respond dynamically as new targets or new diagnostics are required. The ignition target consists of an ablator capsule and fill tube, hohlraum, laser entrance hole (LEH) inserts, thin polymer membranes ('tents'), and the thermal mechanical package (TMP). A silicon cooling arm, thermal shell, diagnostic band, windows, heaters and sensors, tamping gas line, and wiring harness comprise the TMP. Production of the capsule starts with a very smooth, very spherical plastic mandrel, the form on which the silicon-doped plastic (CH) is deposited as the mandrel is rolled or rotated. Following deposition, the capsules are polished. Because the mandrel is plastic, it can be removed by carefully heating the shell to pyrolyze the plastic, resulting in gases that can escape via permeation through the capsule. HDC [37–39] and beryllium capsules [40–42] are also used and manufactured by similar methods. A five-to-ten micron fill hole is then laser-drilled. A precisely drawn and finished tube of glass is used to make the five-to-ten-micron fill tube that is attached to the capsule for DT gas delivery.

The ignition hohlraum is a gold or gold-coated (originally proposed to prevent oxidation) uranium cylinder designed to couple as much laser energy to the capsule as possible. The layers are sputter-deposited on a precision-machined mandrel, which is etched away after deposition and machining are complete. A flange at the waist of the hohlraum supplies an attachment surface for the thin Formvar 'tent' that positions the capsule.

The LEH inserts are machined as separate parts to allow for various LEH diameters, and the LEH windows are aluminum-coated 500 nm thick polyimide film. The diagnostic band joins and aligns the sub-assembled target halves and provides ports for characterization access. The TMP precisely positions the hohlraum, manages the thermal environment of the hohlraum and capsule, and provides a modular platform for various diagnostic configurations without changes to the remaining components. The TMP consists of two precision-fabricated aluminum shells joined by a precision-fabricated band with cutouts to accommodate various diagnostics requirements. Two silicon cooling arms attached to either end of the TMP assembly conduct heat away from the hohlraum to maintain the required temperature. These are lithographically etched to create a precise heat transfer path that ensures temperature uniformity in the target. Heaters located on the TMP shells are then used to produce a nearly spherical isotherm around the capsule. Tamping gas lines deliver helium gas (or other gasses) to the hohlraum to mitigate the interaction between ablated material from the hohlraum walls, as this interaction can steer and scatter the laser light entering the hohlraum.

Components (hohlraum, capsule, etc) for the various platforms are produced largely by General Atomics Corporation. The time to the target use is critical regarding the quantified gas composition (e.g. due to tritium decay). The formation of frozen DT layers is described in [43–45].

We note that it also possible to attach small amounts of target material (thin foils) to the hohlraum or as a thin layer doped into the capsule shell to study for example neutron induced nuclear reactions utilizing radiochemistry methods (see 2.1.9).

The NAD is based on measuring neutron-induced activation in different materials [51–53]. The neutrons are produced in fusion reactions of DD, DT and THD surrogate shots (see p 3, table 1). For example, the 115In(n, n)115 min reaction measures the D(D, n) neutron yield; both 90Zr(n, 2n)89Zr and 63Cu(n, 2n)62Cu measure the D(T, n) primary (13–15 MeV) (figure 5) yield. Tertiary (>15 MeV) neutron yields from energetically up-scattered and reaction-in-flight neutrons are measured utilizing the 209Bi(n, 4n) and169Tm(n, 3n) reactions [54]. Indium and zirconium activation are used together to determine the DSR along many lines of sight [55]. These calibrated diagnostics provide the absolute yield measurements for the principal D (T, n) α reaction at the NIF independent of any other diagnostic system; in particular, yield measurements from NTOF diagnostics are calibrated to the activation yield.

The primary neutron yield, Y_DT, is calculated from the activity measurement of the sample by Y_DT = A₀/(p_rx · ε_irr), where A₀ is the initial activity product nuclei at t = 0; p_rx is the reaction probability for activation foil per neutron; and ε_irr is the irradiation efficiency (neutron fraction at emission center at subtended solid angle at activation target). The reaction probability per primary neutron, p_rs is calculated as p_rs = N_target/4πR²ʃdn/dE σ(E)dE, where N_target is the number of target nuclei; R is the radius of the target from target center; and n is the number of neutrons emitted at target center.

The specific thin disks of activation material, Zr for activation with 14 MeV neutrons from DT in ⁹⁰Zr(n, 2n)⁸⁹Zr reactions, are located in insertion devices at specific locations both within the chamber and are also distributed at specific locations on the outside of NIF chamber flanges (FNAD) (figure 6) [56, 57]. The distribution of measured neutron yields monitors the absolute fuel ρR and its angular distribution in high yield shots since a larger ρR means less activation in the direction of the corresponding activation target due to greater scattering of the fusion neutrons out of that line-of-sight. An accuracy of better than 5% (absolute) and 1%–2% (relative) for shot-to-shot measurements is required and achieved. Substantial experimental effort is dedicated to determining the symmetry and shape of the burning core in ignition experiments. Anisotropy of the cold fuel in the stagnated assembly is believed to be a primary cause of performance degradations since shell asymmetries suggest inefficient compression. This energy deficit appears in radiation-hydrodynamic simulation studies which highlighted the role of insufficient kinetic energy coupling to the hot core. This limited coupling has observable anisotropic nuclear diagnostic signatures, hence spatially resolved areal density diagnostics are essential to diagnosing and optimizing implosion performance.

The neutron spectrum, dn/dE (θ, ϕ), seen at a particular FNAD location, is the burning core neutron production integrated along the attenuation path r through a medium with variable density. The material properties of the medium are also a function of radial path, consisting of burning DT gas, cold compressed DT fuel, and ablator material. Although the ablator mass distribution during the shell implosion phase is measured in surrogate experiments, core emission from a high compression implosion limits the minimum radius at which backlit shell radiography is useful in determining the shape and density of the cold fuel. Path integrals from the hot spot to TC wall can be substantially different between FNAD lines of sight, depending strongly on the shape of the cold fuel. Thus the FNAD signatures offer valuable insight into the important stagnation regime. As an adjunct to the FNAD assembled fuel signatures, a line-of-sight (equatorial) view of both the hot core shape and the cold DT fuel are directly measured by neutron imaging (NI). While artifacts of cold fuel geometry may be seen in the down-scattered neutron image, metrology of two three-dimensional objects (burning fuel and shell) by means of a single line-of-sight 2D image projection, results in an unconstrained family of solutions. In particular, the kinematically unfavorable large angle neutron scattering makes any cold fuel regions at large angles and distances from the NI LOS virtually invisible. Thus the FNAD diagnostic complements and augments the neutron image analysis of the stagnated assembly.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Any one of the specific FNAD locations samples all source neutrons emitted in its direction by the burning core. Those neutrons are attenuated by interstitial cold fuel density integrated along the neutron path length. The solid angular area sampled is then the ratio of the area of emitted neutron beam to the spherical area encompassed at the radius of scattering. In terms of the geometry illustrated in figure 7, assuming circular profiles in the perpendicular dimension, the emitted beam area (π/4) ${{{l}}_{{i}}}^{2}$ divided by the spherical scattering area $4\pi {{r}_{{i}}}^{2}$ gives a sampled solid angle of l²/16r² when l_i ≥ 2r_i and thus the object casts an umbral neutron shadow on the TC wall. The umbral shadow results in a sharp angular gradient in the un-scattered neutron flux and thus proximate FNAD locations show large variations.

Zoom In Zoom Out Reset image size

When l_i < 2r_i the scattering mass no longer obscures the entire hot spot and as such casts only an antumbral shadow on the TC wall. In this case, FNADs over a wide area would see a similar small depression in source strength caused by the small obscuring object.

A bulk center-of-mass velocity of the burning core may also Doppler shift the neutron spectrum. This effect changes the spectrum-averaged activation cross section, as the velocity vector projection changes with observation position.

Studies on warm, gas-filled exploding pusher target implosions reported significant anisotropy attributed to target alignment [57], target mounting [58], capsule defects [59], and non-uniform illumination [60]. For example, the effect of dropping two of the 48 NIF laser quads leads to substantial asymmetry in density and implosion velocity without precise redistribution of the remaining laser energy.

The FNAD activation data and analysis can quantify the effects of bulk velocity within the hot core, especially with respect to the subsequent neutron yield anisotropy. An example of this analysis is presented in figure 8. The relative activation ratios at the different sites previously displayed in figure 6 are plotted with a superimposed low-order spherical harmonic fit to highlight the anisotropic activation distribution for shot N120217-001. Since this shot had inconsequential areal density by design, the observed asymmetries may be ascribed to bulk velocity gradients in the burning plasma. Both a low-order north–south feature and a large amount of high-mode structure are apparent.

Zoom In Zoom Out Reset image size

An analysis that assigns a coherent bulk neutron velocity of 100 km s⁻¹ results in 3.16% additional induced activity in the direction of the velocity, 2.77% of which is due to the change in cross section over the energy shift. The remaining 0.39% is due to the increased number of neutrons emitted in the velocity direction in the lab frame. By detecting the coherent shift in activation around the TC, a single three-component velocity vector may be fit.

Removing the velocity effect on the reported activation ratios (both fluid flow and Doppler shift) leaves just the effect of scattering mass in each FNAD LOS.

An FNADs system based on a lanthanum bromide scintillators and compact gamma spectrometers installed in the target bay is being developed to provide immediate postshot readout of yield [58]. This system will have additional capability to measure multiple DT neutron reactions and other neutron source terms, such as TT and DD fusion reaction neutrons [59, 60].

In ICF experiments with D₂, DT or THD targets the nToF diagnostics measure the neutron yield as a function of neutron energy by recording detector signal strength over time using the known flight path length to deduce the neutron energy. There are several nToF detectors at different lines-of-sight on the NIF whose purpose is to measure different aspects of the neutron spectrum with high accuracy. All nToF detectors have their signals recorded with high bandwidth multi-channel digital oscilloscopes. It is important to note that in contrast to most beam target experiments these nToF detectors operate in current mode looking at the combined signal of many neutrons in a single flash. Most nToF detectors utilize scintillation detectors coupled to PMT and/or PD, however a few detectors use CVD diamonds. Those CVDs were chosen for use in measurements where sensitivity was of lower importance and a fast time response was needed. However, it has been discovered that the IRF of most CVDs in use is not stable; current efforts are therefore exclusively dedicated to the scintillation detectors. Two legacy nToF detectors are at 4.5 m from TCC, which are using fast scintillation detectors. Due to the short flight path those detectors can only be used for measuring the overall neutron yield; one of them is dedicated for low yield shots (10⁷–10¹³ neutrons) while the other covers the yield region from 10¹³ to 10¹⁵ neutrons.

For measurements of spectral information such as ion temperature and the ratio of down scattered neutrons relative to the primary neutron yield down-scattered ratio (DSR), four especially developed detectors are being employed at a distance of about 20 m (the exact distance varies by detector) from TCC. Those detectors use an LLNL developed scintillator [61], which has an extremely low afterglow and is incorporated in a low mass housing to reduce background from neutron scattering. Figure 9 shows a schematic of the detector housing with its mounting and the scintillator inside. Due to the above-mentioned improvements, this detector allows the measurement of the down-scattered neutron spectrum with high accuracy.

Zoom In Zoom Out Reset image size

To achieve the highest possible signal to noise ratio, each PMT/PD signal is split up and recorded on 4 scope channels with a different sensitivity. Two of these channels are stitched together to improve the dynamic range of the data acquisition needed for DSR measurement.

The detectors were designed to obtain spectral information on shots with neutron yields as low as 10¹⁰. This range is obtained by using PMTs with different gains with different light attenuators in front of them. The IRF of these detectors is 4.4 ns FWHM giving them a neutron energy resolution of about 300 keV for 14 MeV neutrons.

The neutron spectrum is characterized by a certain number of key quantities (figure 10(a)): for a DT shot those are the primary neutron yield, defined as the number of neutrons in the 13–15 MeV energy region and the second central moment of the primary neutron peak and the DSR defined as the ratio of neutron yields from 10 to 12 MeV relative to those from 13 to 15 MeV. If there are no motion effects (e.g. averaged velocity distributions in hot spot [62]) broadening the neutron peak then the second central moment is directly proportional to the neutron yield weighted average ion temperature of the plasma [63, 64]. The DSR provides valuable information about the fuel ρR achieved in the implosion. In addition, gated PMTs measure the peak of the DD neutrons emitted from the DT capsule (see figure 7(b)): using a model for the background under this peak the DD neutron yield and ion temperature can be determined.

Zoom In Zoom Out Reset image size

Neutron yield measurements with nToF detectors are only relative; to obtain an absolute yield a calibration to an appropriate NAD is needed. The accuracy of all these measurements is as follows: 7% for the neutron yield coming from the NAD calibration, 300 eV for ion temperature and 10% for the DSR measurement. The nToF spectra are analyzed by using a forward fit of the data, which involves parameterizing the neutron spectrum with a model, taking into account the detector sensitivity and the beam line attenuation and convolving it with the IRF; an example fit to the data is shown in figure 10.

The fit function has the form

$$\begin{eqnarray*}&&f(t)=I(E(t))s(E(t))a(E(t)){\rm{d}}E/{\rm{d}}t\otimes R(t),\end{eqnarray*}$$

where I(E) denotes the model of the neutron spectrum, s(E) is the sensitivity of the detector to the given neutron energy, a(E) is the beam line attenuation factor, E(t) is the neutron energy as a function of time-of-flight, dE/dt converts the energy spectrum into time-of-flight space, R(t) is the instrument response. The neutron energy as a function of time-of-flight can be calculated as E(t) = m_nc²(γ − 1) using γ = (1 − β²)^−1/2 and β = t/t_γ,with t_γ being the time-of-flight of a photon and m_n the mass of the neutron. This relationship determines the neutron energy change dE/dt = m_nc²γ³β³/t_γ. The determination of the instrument response R(t), the scintillator sensitivity s(E), the beam line attenuation a(E), as well as the models used for I(E) and their application in the data analysis are explained in [65].

The burn-weighted ion temperature of the fuel in ICF targets can be deduced from the neutron energy spectrum for DD yields >10¹⁰ and for DT yields of >10¹¹ (if bulk fluid motions are negligible). It is should be noted that an extension of the flight pass (e.g. to 100 m) makes a correction of the energy spectrum (peak width) at the primary peak for the instrumental time resolution (IRF) negligible.

To convert nToF into neutron energy the time offsets for the signal must be known accurately. The time of the scope trace can be interpreted as a sum of several contributions: t = t_E + t_Bang + t_LPOM + t_fidu + t₀ where t_E is the time-of-flight of the neutron from TCC to the detector due to its initial energy, t_Bang is the bang time, i.e. the time it takes from the beginning of the laser pulse to the nuclear burn of the fuel, t_LPOM is the time between the 0 time of Laser Program Operations Management (LPOM) and the actual time when the laser beams hit the target, t_fidu is the position of the timing fiducial on the scope trace and t₀ is a time offset that has to be determined from a x-ray timing shot. Comparing the t₀ for different timing shots the uncertainty of t₀ has been established as σ ≈ 30 ps. An x-ray timing shot produces x-ray emission by illuminating a gold disk with an 88 ps laser pulse. Besides establishing detector timing these shots are also used for constructing the IRF of the nToF detectors. Since the width of 88 ps is negligibly small compared to the response of the instrument the signal recorded on a timing shot serves as the IRF for x-rays. To convert it to an IRF for neutrons the neutron transit time through the 2.5 cm thick scintillation crystal must be taken into account. This is done by using an MCNP simulation to determine the energy deposition of the neutrons over time and convolving it with the x-ray IRF. This procedure must be separately repeated for DD (2.45 MeV) neutrons and DT (14.05 MeV) neutrons since their transit time is significantly different.

In addition to neutron yield and ion temperature, the model described by Ballabio et al [64] relates the mean energy of the neutron peak to the ion temperature; this relationship thus identifies the offset of the peak position as a fluid velocity. This relationship holds independently for both the DD and the DT neutron emission peaks. The uncertainty in this velocity measurement is approximately 15 km s⁻¹, but it is typically observed that the velocity obtained from the DD peak is systematically 10–25 km s⁻¹ lower than the one from the DT peak. One interpretation of this observation is that the ion temperature deduced from the width of the peaks is too high since collective motion can also broaden the peak. Another difference is that the DD peak is much more strongly affected by down-scatter: on a shot with a DSR of 5% more than 50% of the DD neutrons scatter out of the peak region, the exact impact on how this affects the results of the data analysis is not yet understood [66].

The nToF diagnostic is complementary to other nuclear diagnostics (e.g. radiochemistry, NI, neutron activation and γ-ray reaction history) and some of the non-nuclear diagnostics thus providing some redundancy. No single diagnostic alone will provide unambiguous information which allows to infer progress on the ignition experiments. A correlated data analysis is needed to evaluate the ignition or failure conditions.

Based on the magnet-based neutron spectrometer developed for Magnetic Confinement Fusion [67] and installed on the Joint European Torus in Oxford, UK [68], the MRS concept for ICF applications [69, 70] was first demonstrated in 2008 at the OMEGA laser facility. Another MRS system, described in detail in [71–74], was subsequently developed for the NIF (figure 11). This system provides time-integrated information about the neutron spectrum (in the range of 4–23 MeV) through measurements of recoil deuterons produced by neutrons incident on a deuterated polyethylene (CD) foil located 26 cm from the implosion. A small fraction of the recoil deuterons produced in the forward direction is selected by a 20 cm² aperture positioned 570 cm from the foil. These recoil deuterons are momentum analyzed and focused by a magnet onto an array of CR-39 detectors positioned at the focal plane of the spectrometer. Both the magnet and detector are located on the outside of the NIF TC wall. From the measured recoil-deuteron spectrum, the neutron spectrum can be reconstructed. The detector array consists of nine single-use CR-39 detectors, which are processed post-shot in a dedicated etch- and scan laboratory. These types of detectors have the advantages of being 100% efficient for detecting the recoil deuterons and very insensitive to background neutrons, gamma rays and EMP, which allows for measurements with excellent signal-to-background (S/B).

Zoom In Zoom Out Reset image size

MRS routinely measures the neutron spectrum from cryogenically layered DT target implosions [73], from which DT yield, areal density (ρR) and ion temperature are determined (ρR is inferred from the ratio of down-scattered neutrons (10–12 MeV) to primary neutrons (13–15 MeV)). MRS also measures directional flow velocities [74], neutron yields from the T + T reaction [76] and CH ablator areal density in symmetry-capsule implosions with low DT fuel-ρR [75]. MRS is ab initio calibrated and is one of only two independent ways of measuring absolute DT yields on the NIF; the other one is the NAD technique (described later in this section).

An optimal balance between detection efficiency and resolution for a set application can be obtained by varying the thickness and size of the CD foil. For lower yields, thicker foils are used to optimize signal statistics at the expense of resolution. At higher yields, thinner foils are used for better resolution and to avoid track overlap on the CR-39 at high signal levels [77]. As the main sources of background are due to unscattered neutrons and neutrons scattered off structures near the MRS CR-39 detector system, the signal to background is reduced when thinner foils are used.

To allow for high-fidelity analysis at low signal levels with high relative background, such as for the DSR measurements, a dedicated CR-39 processing technique, the so-called coincidence counting technique [78], has been developed. This involves processing the CR-39 detectors in two steps, eliminating background by only accepting as signal events that occur in the same location in the two layers, and allows up to ∼2 orders of magnitude improvement in S/B.

The systematic uncertainty in the measured absolute yield comes from the uncertainty in the MRS geometry and inaccuracy of the CD foil characterization. For some configurations, the systematic uncertainty is as low as 4%. The systematic uncertainty in the measured DSR is set only by the relative uncertainty in the n, D elastic cross section at 10–12 MeV and 13–15 MeV and is less than 5%.

MRS can also be operated in charged-particle mode to measure the energy spectra of protons (E_p ∼ 6–40 MeV) or deuterons (E_d ∼ 2.5–20 MeV) directly from the implosion with an energy resolution of ∼0.15 MeV. In this case, MRS is fielded without the CD foil, and has a detection efficiency of 4.3 × 10⁻⁶ set by the solid angle covered by the magnet aperture. The dynamic range for MRS in this mode depends on the shape of the probed spectrum; peaked spectra can be measured from ∼10⁸ upwards with enough signal to pick out the tracks from the intrinsic CR-39 background noise and broad spectra covering several MeV could be measured up to at least 10¹¹ without running into track overlap on the CR-39.

The NIF NIS was installed in 2010, commissioned in 2011 and is presently being used to provide data on the size and shape of the fusion hotspot and the surrounding cold fuel for ICF implosions [79]. This system is comprised of three basic pieces: an aperture array is used to form the neutron images, a detector system is used to measure the neutron flux passing through the pinhole array, and reconstruction algorithms have been developed to extract the neutron source distributions from the recorded images. The application of this technique to measure the underlying burning fuel distribution has been developed in the past decade and now implemented at NIF [80–84].

This detector system [85] can collect two images, which are fast gated and independently timed. Because the detector array is positioned 28 m from the neutron source, the neutron arrival time at the detector is correlated to the neutron energies, allowing measurement of the neutron source distributions from two energy ranges by gating the detectors at two different times. Typically, one detector is gated to view the 14 MeV neutrons (13–17 MeV), which are generated from DT fusion processes, while the second detector is gated to measure the source distribution of lower energy neutrons, typically in the range from 6 to 12 MeV. These lower energy neutrons are predominantly DT fusion neutrons that have scattered in the surrounding cold fuel and, therefore, provide information on the distribution of cold fuel surrounding the hotspot. By measuring these sources along the same LOS the images can be co-registered allowing the two images to be placed relative to each other.

Because of the long mean free path of neutrons within the aperture material, the NI system forms images using extended apertures or 'pinholes' in materials with short interaction lengths. The aperture array currently in use at NIF is composed of triangular pinholes and mini-penumbral apertures machined along the 20 cm length of wedged gold layers. Gold has an interaction length of ∼3 cm for 14 MeV neutrons. The pinholes have an equilateral, triangular cross-section and taper from a height of 0–210 μm in the direction of neutron propagation. Each tapered pinhole provides a field of view of approximately 200 μm at 26.5 cm from the entrance face of the aperture array. A total of 20 pinholes have been precision machined into four tapered gold layers, which are stacked to form the aperture array. The mini-penumbral apertures are double-tapered cylinders with a 273 μm input aperture diameter, a 300 μm waist diameter, and a 437 μm exit aperture diameter. Three apertures were formed by machining three half cylinders into each of two gold slabs and precision stacking the two layers to form the penumbral apertures. Each aperture has a field of view of 200 μm. At the present operating position, the field of view of the ensemble of pinholes to <0.7 mm horizontally and <0.5 mm vertically, as each pinhole now points at a slightly different location at the source.

The image recording system is located 28 m from TCC with two collimators placed along the neutron flight path providing a measurement environment with low neutron backgrounds.

Neutrons that form the images interact with a coherent scintillating fiber array. This fiber array is 160 mm square and is composed of BCF-99-55 scintillating fibers. The round fibers are 250 μm in diameter and 5 cm long. This light is transported within the fibers to the two ends of the fiber array. A turning mirror located at the upstream end of the fiber array directs the scintillation light into an optical lens, which forms an image on a 75 mm micro-channel plate (MCP). This MCP provides signal gain and the fast gating of the image collection system. The output image of the 75 mm channel plate is reduced to 37 mm with a fiber taper that couples to the fiber optic window of a SI-1000 CCD camera.

The light that exits the fiber on the downstream ends is collected into a fiber taper reducing the image from 160 to 75 mm into a second 75 mm MCP. The output of this MCP is collected with a 37 mm coherent fiber 'rope' and directed to a second SI-1000 CCD camera. The detector resolution at the scintillator plane of both image recording systems has been measured to be ∼1.2 mm FWHM, resulting in 15 μm resolution at the source plane after correcting for the magnification.

By injecting a pattern of neutrons into the scintillator and gating both imaging systems at the same time, the two NI systems have been spatially calibrated into the same coordinate system. This allows the primary neutron image to be overlaid on the down scattered neutron image to within ∼5 μm, allowing detailed analysis of the image data correlating neutron emission with the fuel assembly.

Because of the small neutron source sizes and the triangular shape of the neutron pinholes, significant effort has been devoted to the development of analysis procedures to remove imaging distortions. Figure 12 shows a primary and down scattered neutron image from all the pinholes for shot N140304, the highest yielding NIF ICF implosion to date, which generated nearly 10¹⁶ neutrons and figure 13 shows the five pinholes with the field of view overlapping the neutron source.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Several maximum-likelihood estimate techniques have been employed to remove the distortions shown in figure 13. The distortions introduced by neutrons interacting with the aperture housing is calculated and characterized through a four-dimensional matrix. This four-dimensional matrix maps the attenuation properties from each position at the source plane to each position at the detector plane. The two-dimensional source plane position is then mapped into the two-dimensions of the image plane, and these transport properties are characterized within this four-dimensional matrix.

With this formulation, the reconstruction process can be characterized as the inversion of this matrix formula. This is a notoriously difficult problem as this ill-posed inversion is under-constrained. Maximum-likelihood techniques have been formulated for this inversion process [86]. An algorithm which assumes Gaussian noise and employs an adaptive step size has been shown to be the most robust for reconstructions of the underlying sources. Because multiple pinholes each provide an image from a NIF implosion, four-dimensional matrices are calculated for the detailed pointing solution for each pinhole. The maximum-likelihood techniques are used to reconstruct the underlying source that is consistent with each pinhole image. This technique provides significant enhancement over single pinhole techniques as any systematic errors introduced by each pinhole are partially corrected when simultaneously used in an ensemble reconstruction.

Figures 14 and 15 show the reconstructed results of the data shown in figure 13, showing the underlying reconstructed source as well as the 17% contour, which has been fit to Legendre polynomials. The 17% contour is typically reported for NIF experiments, as this contour level should closely agree with the 17% high-energy x-ray contour.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

The NI system at NIF has been installed and commissioned and is providing hotspot and cold-fuel shape information to the National Ignition Campaign for tuning the implosion of the ignition capsules. The large amount of data that has been collected since early 2011 has provided substantial gains in system characterization, data analysis and source reconstruction algorithms. This data continues to be analyzed to further improve system performance and the development of analysis algorithms.

This technique has proven so powerful [87, 88] that an effort is underway to install a second and third NI LOS along the polar and another equatorial LOS at NIF. The combination of this data would provide information on the three-dimensional nature of the implosion system at stagnation.

In the idealized picture of a compressed DT capsule at stagnation, a spherical shell of remaining ablator surrounds a spherical shell of the remaining cold fuel that in turn surrounds a sphere of burning DT fuel. Reality, however, imposes itself during compression in the form of asymmetric drive, native roughness of the capsule, the fill tube, the tents that hold the capsule, and instabilities. These features and effects can cause distortions and mixing and produce asymmetries in the compressed ablator and DT fuel and in the burning hot spot. Since these realities can also limit the yield, understanding the spatial features of the hotspot, cold fuel and ablator throughout the compression and burn is crucial.

While radiography and imaging of self-emission x-rays are the primary methods for observing the symmetry of the capsule during compression [89–91], current radiography methods go blind due to the high capsule density prior to stagnation, and the x-ray images are strongly influenced by the emitting materials and opacity near stagnation. To understand the shape of the burn volume, whether the fuel is burning where it is emitting x-rays, or how the remaining ablator or cold fuel are distributed around the burning core, the 14.1 MeV neutrons that are produced by the burning fuel provide a useful probe.

Direct NI provides an immediate picture of the burning DT source since only a small fraction of the neutrons interacts before escaping the compressed capsule. The fraction of neutrons that does interact with the compressed capsule before exiting is also useful. Neutrons that down scatter from the cold DT fuel and lose energy can be separated from the 14.1 MeV neutrons by time of flight and imaged to obtain information on the density of the remaining cold fuel. Other neutrons can interact with the carbon in plastic ablators and scatter, producing a 4.43 MeV γ-ray through the reaction:

$$\begin{eqnarray*}&&{C}^{12}(n,\,n^{\prime} ){C}^{12* }\to {C}^{12}+\gamma (4.43\,{\rm{MeV}}).\end{eqnarray*}$$

Imaging those γ-rays then provides information on the density of the remaining ablator.

The existing NIS on the 90–315 LOS [92], which is described above, provides direct images of the fusion hot spot by imaging the 14.1 MeV neutrons and images of the cold fuel by imaging of the down-scattered neutrons. By sacrificing one of these images and gating one arm of the NIS system on the photon arrival, this system has also demonstrated imaging of γ-rays from the remaining ablator. Due to the inherently three-dimensional nature of the implosion, single views of the 14.1 MeV neutrons, the down-scattered neutrons, or the γ-rays are insufficient to understand the source volumes. For imaging the burn volume, simulations have shown that two views (figure 16)—one equatorial and one polar—reduce the assumptions required for 3D reconstruction and that three near-orthogonal views eliminate the need for symmetry assumptions and eliminate most reconstruction artifacts. Reconstructions from γ-ray images should perform similarly. For the down-scattered images, where the kinematics of the scattering cause images to sample strongly in the direction of the LOS, one would expect six views to be required along three near-orthogonal axes.

Zoom In Zoom Out Reset image size

Moving beyond a single view with two images will require additional lines of sight, each of which has integrated measurements of the 14.1 MeV neutron, down-scattered neutron and γ-ray images, preferably with x-ray imaging on the same or nearby lines of sight.

As part of the move to multiple lines of sight, the NIF has installed an energy-integrated neutron imager on the 5.25–225 LOS. This imager uses a detector stack of four pair of 2 mm polyethylene sheets followed by image plates. The detector relies on (n, p) reactions in the high-density polyethylene to produce protons that interact with the image plate to record an image [93]. While it has lower efficiency than the 50 mm thick BCF-99-55 scintillating fiber array used on 90–315, this imaging stack has excellent resolution (∼650 μm FWHM point spread function versus 1.1 mm for the BCF-99-55 fiber array) and allows imaging on an 8.2 m LOS. Adding energy resolution or γ-ray imaging on this LOS will require improved scintillators. An additional equatorial LOS is also planned to allow the required set of near orthogonal measurements.

In ICF experiments nuclear reaction products are produced over the burn time in the region of the burning fuel. The MeV neutrons and gamma rays produced probe the burning fuel uniquely since they escape from the capsule-centered burn region relatively undistorted. The nuclear bang time—the time interval from onset of the laser pulse to the maximum of neutron and fusion gamma emission—is measured accurately via the time resolved detection of neutrons and gamma rays. The x-ray signals provide a bang time as well but are subject to opacity issues and competing signals from other pre-burn sources such as supra-thermal electron Bremsstrahlung produced by the hohlraum plasma. The time history of the nuclear DT fusion reaction (burn), the evolution of the ignition to peak burn and beyond, is most promptly determined by monitoring the DT fusion gammas that have primary peak energy at 16.7 MeV with high time resolution. The gammas have no time-of-flight dispersion along their flight path; the dynamic range, however, is yield limited because of the 4 × 10⁻⁵ branching ratio of DT fusion gammas to the 14.1 MeV neutron emission [94].

The GRH diagnostic employs Cherenkov counters running in current mode in four separate gas cells with variable pressures to adjust the Cherenkov threshold on Compton-scattered electrons in the active detector volume [95]. The chosen thresholds typically correspond to γ-ray energies of 10, 8, 4.5 and 2.9 MeV. The 10 and 8 MeV thresholds measure the fusion γ-rays at 16.7 MeV. Gamma rays produced from neutron interactions with materials surrounding the target dominate the 4.5 MeV threshold detector, while the 2.9 MeV signal divides about evenly between ¹²C (n; n'γ_4.4) and the neutron-induced γ-ray background.

In the GRH itself, gammas are converted into relativistic electrons, primarily by Compton scattering in a low-Z convertor plate. The electrons then produce 'Cherenkov light' in the GRH gas medium at a given index of refraction (1/n_R ∼ v/c) which is dependent upon the gas pressure. Different conditions, such as gas pressures of 200–40 psi for CO₂ or SF₆, chosen for the GRH medium, result in different energy thresholds for the gamma rays. The blue Cherenkov light is transported via a mirror system to a photomultiplier tube (or possibly streaked system in the future) that includes a Mach–Zender interferometer transmission system to transfer the signal to a fast oscilloscope. The multiple cell GRH system at the TC wall measures the energy-selected gammas (3–20 MeV) with a time resolution of approximately 50 ps. The DT gammas can be effectively measured over a neutron-yield range from 10¹⁴ to 10¹⁹.

Estimates of the detector signal voltage follow from the various conversion factors and detector efficiencies. For example, a relatively low yield of 10¹⁴ neutrons implies production of 4.0 × 10⁹ DT-fusion gamma rays using the measured branching ratio of 4.0 × 10⁻⁵; the collection solid angle of 3 × 10⁻⁵ steradians thus captures approximately 10⁵ gamma rays. These captured photons undergo Compton conversion at the 0.1% level; the resulting 10² Compton electrons produce 10 Cherenkov photons per electron, or approximately 10³ total Cherenkov photons. Amplification of 10⁵ by the PMT produces 10⁸ signal electrons within 100 ps for a current of 10¹⁸ electrons s⁻¹ or 0.1 Amps; the measured 50 Ω resistance ensures a 5 V signal. Obviously, the measurement becomes increasingly accurate at higher neutron yields.

For low yield shots, with alternate gas fills such as THD, D³He, or HT, a GCD, using a Cassegrain mirror system in a re-entrant tube extending into the TC 4 m from the target is being developed by LANL. The ability to insert the GCD into the evacuated TC on a DIM or TANDM mount is also being pursued to extend the lower neutron yield limit to 10¹¹. A time resolution of less than 30 ps for the bang time, burn width and reaction history measurement is anticipated. Additionally, an accuracy of 20%–25% (absolute) and 5%–10% from shot-to-shot variation at a SNR of about 30 is desirable. Tests and calibrations performed at the HIGS FEL facility at Duke University have demonstrated that these requirements are achievable.

As a more speculative goal, gamma ray emission caused by inelastic (n, n'γ) reactions with different elements present in the ablator shell or hohlraum material might be used as an additional reaction history diagnostic. In the case of a plastic (hydrocarbon) capsule, simulations show intense 4.4 MeV ¹²C gamma rays and 6.2 MeV gamma rays from oxygen contamination, as displayed in figure 17(a). The cross sections are large for down-scattered (1–12 MeV) and thermalized (<1 MeV) neutrons. Sufficiently short time resolution may enable direct detection of time-shifted emission from hohlraum material. For hard x-ray emission, arising from self-emission during burn, simulations predict a time delay with respect to the fusion gamma emission; this time shift provides a measure of the ion temperature thermal equilibration process.

Zoom In Zoom Out Reset image size

As an example of typical data reduction, GRH spectra in figure 17(b) displays a least squares data fit, which decomposes the 2.9 MeV channel data into a ¹²C(n, n'γ_4.4) signal and two backgrounds. Constraints are placed on each background contribution. Calibration with an exploding pusher target with negligible ablator and DT areal densities and the high-foot target holder determine the neutron-induced background for a given neutron yield. A contribution of less than 5% background from the T(D, γ) reaction is also taken into account. The T(D, γ) yield, extracted from the two highest thresholds, combined with its γ-ray spectrum and the detector response, determines the 2.9 MeV background. Calibration of the detector for ¹²C(n, n'γ_4.4) has been performed in situ with an DT exploding pusher and a thin (1 cm) graphite disk. A conversion factor that transforms the time-integrated γ-ray signal from 2.9 MeV data into a total number of ¹²C(n, n'γ_4.4) reactions was thus determined from this calibration.

The ¹²C gamma-ray measurements are used to deduce the areal density of the remaining hydrocarbon ablator in the stagnated fuel assembly [98]. The ¹²C (n, n'γ_4.4) yield on a specific shot is determined by integrating the fitted signal and multiplying the result with the measured conversion factor. The ablator areal density measurement depends on the ratio of 4.4 MeV rays to neutrons. While a calibration has been performed with a known ¹²C areal density, the ablator areal density accuracy depends on neutron spectral analysis between the calibration shot and the measurement in question. For high compression implosions, the DT areal density modifies the birth neutron spectrum by inelastic scattering processes; this component requires a correction to the ¹²C (n, n'γ_4.4) reactions at energies less than 14 MeV. Folding the down-scattered spectrum with the ¹²C (n, n'γ_4.4) cross section establishes a correction of approximately 20% to the measured carbon ρR. Uncertainties from the down-scatter correction and detector calibration limit the accuracy of the ρR measurements to ±20%.

Another interesting diagnostic signature may be accessed for lower yield shots such as TH (10% D)-filled targets with neutron yields at 10¹². In this case, 20 MeV γ-rays from up-scattered protons in t + p fusion reactions can be observed. These gamma rays give a clean, essentially background free signal and provide a sensitive ρR fuel diagnostic proportional to (ρR)². This γ-ray diagnostic is complementary to some of the radiochemistry diagnostics (see comment in section 3.2.6).

Extending this general approach to energy resolved γ-ray spectroscopy in the picosecond time domain of fusion reactions or of excited nuclei in HED environments is desirable but difficult. The γ-rays with energies less than 1.5 MeV can be analyzed with high-energy resolution by means of crystal spectrometers combined with fast scintillator based detector arrays; this requires high γ-ray fluxes. For γ-rays with energies up to about 20 MeV large fast single event scintillator-based detector arrays are a possible solution. A gamma spectrum emitted from the NIF target capsule can be energy-resolved by utilizing a one-event, macro 'pixelated', detector. Such a single event detector array must be located at a large distance (e.g. 100 m) from the NIF TC; this deployment would be technically challenging.

The current GRH diagnostic is limited in both time response and energy discrimination by its large stand-off distance (6 m) to the TCC, Qualification of the LPI x-ray and prompt hohlraum-gamma background for a new GCD (GCD-3) at the Omega facility has been completed. This assembly will be mounted in a shielded well at 3.9 m, thus reducing the stand-off distance. The relative configurations of the GRH and the well-mounted GCD are displayed in figure 18. An expanded view of the well-mounted diagnostic is also provided in figure 18. An immediate improvement in the measured gamma-ray temporal response will be achieved by using a photodiode detector, lowering the present 100 ps GRH time resolution to 60 ps. In the next deployment phase, pulse dilation technology will be introduced to further shorten the temporal resolution to 10 ps for neutron yields above 10¹⁵. A third phase in the deployment of a GCD ('super GCD') has been proposed to extend the effective neutron yield range to include very low yield (10¹¹) implosions by moving the detector to 20 cm from TCC. Substantially greater insight into ICF implosion dynamics will be obtained with these improvements. Outstanding issues such as residual kinetic energy in the fuel assembly, thermonuclear burn truncation due to material mixing, and the quantification of alpha-particle heating can be directly addressed.

Zoom In Zoom Out Reset image size

Spectra of protons emitted by inertial fusion implosions are measured with WRF proton spectrometers [100–103] on both OMEGA and the NIF. The protons are incident upon a wedged aluminum filter, with thickness varying from ∼150 μm to ∼2 mm. Transmitted protons are detected with CR-39. Within a certain energy range, CR-39 can accurately measure proton energies via the track size, and by accounting for the known thickness of aluminum transited the incident proton energy is inferred [104, 105]. A spectrum is a histogram of all tracks detected on the CR-39. The WRF can measure the proton spectrum between ∼4 and 20 MeV, and for yields ≥10 on NIF.

In fusion research, 'surrogate' CH-shell implosions filled with D³He gas are used as 'mirror reaction' in place of the cryogenic DT ignition targets for measurements of symmetry, implosion velocity, and other studies. In these implosions, the D + ³He fusion reaction generates an energetic proton at 14.7 MeV. Spectroscopy of protons produced at the shock-bang time, while the dense shell is still imploding, has been used to study the implosion shock dynamics [104] and in-flight symmetry [105].

For fundamental science, the WRFs can measure proton spectra from reactions relevant to stellar nucleosynthesis and few-body nuclear physics [106]. The ³He + ³He fusion reaction, the final step of the solar proton-proton I chain, produces protons up to 10.8 MeV, which can be measured using the WRF. The T + ³He reaction, a complementary 6-nucleon system, produces protons with energies up to 12.1 MeV. The spectral shape for these reactions can be used to constrain phenomenological and ab initio nuclear theory of few-body charged-particle reactions, a topic of current research.

The time of peak nuclear production (nuclear-bang time) provides a valuable comparison between ICF implosions and simulations. The PTOF diagnostic was implemented to measure nuclear-bang times in NIF implosions using DT, D₂, D³He, and T₂ fuel [107].

The active element of the PTOF detector is a 200 μm thick, 10 mm diameter circular CVD-diamond wafer with gold electrodes deposited on either side. This wafer is held in a brass housing, positioned approximately 50 cm from TCC by the same hardware that positions the WRFs on the DIM (at chamber location 90, 78), and the rear surface is biased at high voltage (typically −250 V). High energy neutrons and protons produced by fusion reactions in the implosion excite electron-hole pairs when they transit the wafer, which are then swept out by the bias field, producing a current impulse. These signals are recorded as a voltage impulse by a transient digitizer and one channel of a digitizer. The timing of the observed signal trace on both oscilloscopes is absolutely calibrated with respect to the NIF optical fiducial system on periodic x-ray impulse shots, allowing the precise inference of the nuclear-bang time for an observed particle signal. Signals are forward-fit with a particle source function, which is constructed using the observed particle spectrum Doppler-broadened and convolved with the IRF measured on x-ray impulse shots. The particle time-of-flight is subtracted from the signal detection time to obtain a precise bang-time. Analysis of the primary imaging data from DIM (90, 78) provides accurate knowledge of the detector distance from TCC; total uncertainty in the inferred bang times is generally less than ±100 ps.

Due to its proximity to TCC, PTOF is uniquely capable of recording accurate nuclear bang-times using several nuclear products and at very low yields: DT-neutrons (14.1 MeV) with yields above 1.5 × 10¹⁰, DD-neutrons (2.45 MeV) with yields above 3 × 10¹⁰, and D³He protons (initially 14.7 MeV) with yields above 1 × 10⁷. PTOF is the only diagnostic capable of measuring bang-time using DD-neutrons, which are produced in surrogate implosions with D₂ or D³He fuel. Uncertainty of neutron bang times is dominated by uncertainty in the detector positioning, and is typically ±41 ps for DT-neutrons and ±62 ps for DD neutrons. Unlike neutrons, protons are substantially downshifted from their birth energy by ranging in the fuel, shell, and hohlraum, such that uncertainty of proton bang times is dominated by the proton mean energy as measured by the WRFs, and is typically ±90 ps. To date, PTOF has recorded accurate bang times on over 200 experiments at the NIF.

PTOF has demonstrated the simultaneous measurement of DD-neutrons produced at compression-bang time (near the time of peak fuel compression) and the D³He-protons produced at shock-bang time (0.5–1 ns prior to peak compression, near the time of peak implosion velocity; see 'WRF' section for more information) on several near-vacuum hohlraum (NVH) implosions with D³He-gas target fill; see figure 19. These simultaneous measurements of the shock- and compression-bang time differential in indirectly-driven implosions provide uniquely valuable information on the shock propagation and the deceleration phase. Comparison of these results to simulations indicates that the fuel assembly in these subscale, high-adiabat implosions is well-understood. In contrast, a comparison of the ρR at shock-bang to radiography data suggests that the shock-bang time is not well understood in full-scale D³He-gas filled implosions surrogate to cryogenic experiments [108]. Such full-scale gas-filled hohlraum implosions produce much larger x-ray backgrounds than the subscale NVH implosions, such that the x-ray filtering required for the PTOF detector is too thick for protons to reach the detector and PTOF is unable to measure the shock-bang time using D³He-protons directly.

Zoom In Zoom Out Reset image size

The MagPTOF upgrade to the PTOF diagnostic was implemented to measure the shock- and compression-bang time on full-scale implosions using the D³He-protons and DD-neutrons, respectively [104]. The design incorporates thick (1–4 cm) tungsten filtering between the PTOF detector and TCC to reduce the x-ray background by a factor of 1000x, and uses a fixed dipole magnet with peak field strength of 1 T to bend protons with energy between 6 and 16 MeV around the shielding and onto the PTOF detector. The MagPTOF has been implemented as a ride-along diagnostic on DIM (90, 78), utilizing the same detectors, cables, and recording systems as the PTOF.

The MagPTOF system incorporates a piece of CR-39 nuclear track detector surrounding the PTOF detector, allowing it to be fielded as a charged particle spectrometer. When the magnet is fielded with a slit aperture, incident charged particles are deflected onto the CR-39 with their final recorded position a function of the particle energy. The system can record protons in the energy range 2–4 MeV with a resolution of ±100 keV.

This unique capability will be valuable for spectroscopic measurements of 3.0 MeV protons from the DD-fusion reaction for both low-ρR implosions and laboratory astrophysics experiments, such as collisionless counter-streaming plasma experiments to study the Weibel instability [109].

The NIF RC diagnostics are based on nuclear reactions of neutrons or charged particles with detector nuclei that are present in either the target capsule or the surrounding hohlraum assembly.

Following a NIF target shot, the resultant debris products, either gaseous or solid, are collected from the chamber and analyzed via radiation detection or mass spectrometry to determine the concentration of reaction products that were produced in the capsule and hohlraum. The number of reaction products is related to the particle flux produced during the capsule implosion and is used either as a diagnostic of the capsule performance, or as a way to measure reaction cross sections.

Gaseous products are collected via the RAGS system (figure 20) [6]. The gas load from the TC is diverted through a series of turbo pumps to the first stage of RAGS where the gas is filtered to remove unwanted species and reactive gases. This procedure leaves primarily a noble gas fraction for further purification and collection. The system can be set to collect xenon or both krypton and xenon by adjustment of the cryotrap temperature. Radioactive isotopes are first analyzed on the cryotrap via in situ gamma-ray spectroscopy using a HPGe gamma-ray detector positioned above it. The gases can be subsequently transferred and frozen into a removable sample bottle, which is then subsequently analyzed using high-efficiency HPGe gamma-ray detectors or noble gas mass spectrometry. In order to determine the total efficiency for collecting reaction products from the NIF chamber, a small amount of either stable or radioactive tracer gas is injected into the chamber immediately following the shot and pumped with the gaseous products into the RAGS system. The total collection efficiency is determined by comparing the number of tracer atoms in the final gas fraction to the number injected during the shot. This efficiency is then used to extract the total number of radioactive gas atoms that were produced during the shot.

Zoom In Zoom Out Reset image size

Collection of solid debris is performed with the solid radiochemistry (SRC) diagnostic (figure 21) [110]. The SRC consists of 2 inch diameter discs of either pure metals (tantalum or vanadium) or carbon foils mounted in holders on several DIMs [111]. Up to four discs can be fielded at 50 cm from TCC on each of three DIMs—two on the equator at coordinates 90–78 and 90–315 and one on the north pole at 0–0. Each disc has an active surface area that covers 4 × 10⁻⁴ of the 4π solid angle. In addition, a large area detector (vast area detector for experimental radiochemistry, VADER) covering about six times more solid angle than the 2 inch discs can be fielded on DIM 90–78 for increased collection efficiency [112]. Following a NIF shot, the DIMs are retracted from the NIF TC, and the SRC discs or large area detector are removed and transferred for radiation counting via high-efficiency gamma-ray spectroscopy. Depending on the reaction of interest, SRC samples may be first chemically processed to separate the material of interest from other interfering background products. Non-radioactive species are determined via mass spectrometry. Collection efficiency is determined relative to the total number of ¹⁹⁶Au atoms collected from the gold hohlraum.

Zoom In Zoom Out Reset image size

Nuclear reactions occur on materials that are either incorporated into the NIF capsule shell as detector atoms or materials present as part of the surrounding hohlraum assembly. After the lasers are fired, both neutrons and charged particles (protons, deuterons, tritons and alpha particles) are produced in the capsule fuel, which then interact with the surrounding detector material to produce radioactive products. These products are then collected and evaluated using either RAGS or SRC. The number of reaction products produced on any given target material depends on the location of the detector atoms with respect to the source of the particles, the distribution (density gradient) throughout the capsule shell or hohlraum, the energy flux of the particles, and the corresponding reaction cross sections. The DT fuel areal density (ρR) determines the number of primary (14 MeV) and scattered neutrons as well as charged particles (e.g., d, p, α) that induce the reactions that create the radioactive product isotopes. A large fuel ρR results in a higher probability of neutrons being scattered off of the D and T atoms to lower energies. An estimate of the ρR, therefore, can be extracted by measuring the number of radioactive products produced at lower energies in comparison to the 14 MeV neutron yield. The gold in the hohlraum is exposed during a shot to the total neutron spectrum that is produced during the implosion. The activated gold debris is typically collected via SRC and analyzed for products that are created in the ¹⁹⁷Au(n, γ)¹⁹⁸Au and ¹⁹⁷Au(n, 2n)¹⁹⁶Au reactions. The ratio of the neutron capture products, which are produced with larger reaction cross sections at lower neutron energy, to the number of (n, 2n) products, which are sensitive to the 14 MeV neutrons, is correlated to the fraction of neutrons that have been scattered to energies in the 10–12 MeV range (the down scattered ratio, or DSR) [113]. Figure 22 shows the ¹⁹⁸Au/¹⁹⁶Au ratio determined for a series of DT shots at NIF. This ratio provides a diagnostic for the fuel compression since the DSR is directly related to the ρR.

Zoom In Zoom Out Reset image size

When detector atoms are loaded in the NIF capsule, as opposed to the hohlraum, the number of reactions that occur is sensitive to the location of the dopant atoms, whether they are located in a uniform layer or have a density gradient, the symmetry of the implosion, temperature, and mix between the capsule ablator and the fuel. In addition, the range of the incident particles is also a factor. Alpha particles and deuterons have much shorter ranges than neutrons. If tracer materials are loaded at the fuel-shell interface, the observation of charged-particle reactions could provide a useful diagnostic for quantifying the amount of ablator/fuel mix. For example, if ¹²⁷I were loaded in the capsule ablator, the ¹²⁷I(d, 2n)¹²⁷Xe reaction produces a noble gas that can be collected and quantified using RAGS. If ¹²⁴Xe was co-loaded with the ¹²⁷I, the ratio of the (d, 2n) reactions on the iodine to the ¹²⁴Xe(n, 2n)¹²³Xe reaction would provide a symmetry and mix diagnostic where the xenon provides both yield normalization and a measure of collection efficiency. The ratio is integrated over the peak burn (width) while uncertainties in energy and space are eliminated.

The NIF chamber is intrinsically an explosive environment, which makes collection and interpretation of debris more difficult. To interpret the measured number of collected products from loaded materials as a corresponding ρR or mix diagnostic, independent measurements of reaction cross sections are required. In addition, the debris from the capsule and surrounding hohlraum assembly are not uniformly distributed on the SRC collection plates. Therefore, the effects of chemical condensation and fractionation (separation) of volatile species must also be considered in the quantification of reaction products [114]. For the collection of small numbers of atoms, alternative methods of high-efficiency debris collection may need to be considered. This may include development of charged collection plates, or aerosolized gas-jets. Collection prototypes can be developed at accelerator facilities before implementation at NIF.

Besides providing unique mix- and fuel ρR diagnostics, RC can also potentially be used to measure unknown reaction cross sections by introducing detector materials, including radioactive species, into capsule ablators for DT shots. The burn time of a typical NIF capsule is very short, which means that short-lived nuclear states are available as target nuclei for subsequent nuclear reactions. This makes NIF potentially a unique facility for cross section measurements. In a typical accelerator irradiation, the particle flux is too low to be able to create activated species from short-lived states. The number of target atoms required to perform a comparable experiment at NIF is on order of 10¹⁶, which means that even radioactive targets are possible. Another unique feature of NIF is that the contribution from lower-energy neutrons that scatter off of the surrounding chamber materials (so-called room-return neutrons) has been shown to be negligible within the NIF capsule and hohlraum [113]. This means that neutron capture reactions measured on detector materials loaded in NIF capsules will not require large corrections to account for contributions from chamber-scattered neutrons, which is not the case for most neutron facilities.

In order to perform measurements on extremely short-lived species (half-life ∼1 min) a rapid sample extraction system would be needed to remove solid debris samples from the NIF TC for immediate radiation counting following the shot. A proposed design for such a system, the rapid access collection for experimental radiochemistry (RACER), would field a solid debris collector on one of the DIMs at a distance of 50 cm from TCC. Immediately following the shot, the sample would be dropped into a vacuum tube where it would be transported from the TC directly into the basement of the NIF facility. The sample would then be placed on a high-purity germanium detector for gamma-ray spectroscopic measurements. Based on the initial design of the RACER apparatus, the time between collection and counting is on the order of 1 min [113]. This would allow for measurement of short-lived reaction products that would otherwise not be possible based on current sample retrieval times of >2 h post shot.

NIF experimental data are currently limited by inadequate sampling of important spatial and temporal aspects of the critical stagnation phase. Consequently, the existing data must be augmented by simple physical models and more realistic radiation-hydrodynamic simulations to guide post-shot analysis and optimization. Interpretation of a particular observable, or better yet, a collection of observables, requires context provided by physics models of varying sophistication. Due to the complexity and cost of any individual implosion experiment, it is critical to the success of the ICF effort to have reliable, validated models. The most highly developed NIF simulation model to date is that provided by the radiation-hydrodynamic code HYDRA. Its application to experimental analysis will be highlighted below [126, 127].

Before an experiment, pre-shot simulations of the nominal experimental conditions are performed to set our expectations for observations—x-ray brightness, neutron yields, spectral widths, etc. To do this, we simulate not only the implosion physics, but also the detailed output from diagnostics. Our principal numerical diagnostics include synthetic neutron spectra, neutron yield histories, neutron images, and spectrally and time resolved x-ray images, all along multiple lines of sight. Such predicted pre-shot observations are accurate enough to set instrument settings that capture high quality data to avoid mistiming of streak diagnostics or saturation of image and yield diagnostics.

However, no experiment has initial conditions that correspond exactly to the specified pre-shot conditions, and real experimental data almost always diverges from pre-shot predictions to some extent. Target and laser performance can vary within allowed tolerances, and the x-ray drive is difficult to control and monitor. Hence, it is essential to capture the effects of these irreproducible variations. Our detailed metrology of targets and laser delivery give us the necessary measurements to correct our pre-shot simulations for the as-delivered fluctuations encountered during the experiment. These corrected simulations, called post-shot simulations, close a portion of the gap between pre-shot expectations and experimental observations. Yet, differences remain. The simulations are not perfect models of the experiments, and their shortcomings provide us with clues to essential physics that is not captured or is captured incorrectly in our code tools.

Post-shot designers work to tune post-shot observables to the experimental truth. While we know well the time history of laser light delivered to the hohlraum, multiple physics processes alter the temporal and spatial distribution of this energy. Laser light may be scattered by laser–plasma interactions, absorbed laser light is re-emitted as x-rays, and the x-ray drive couples through ablation to capsule hydrodynamics. Anywhere in this conversion chain uncertainties and inadequacies in our models may lead simulations to produce stagnated hot spot conditions that differ from experiments. Designers then make ad hoc adjustments to the driving radiation source to bring the simulated stagnation conditions and associated synthetic diagnostics into consistency with experiments. These adjustments are not intended to provide the physical reasoning for how the laser-hohlraum-capsule coupling ended up the way it did, but rather to highlight the timing and magnitude of driving discrepancies for which we must seek physical explanations and hypotheses. In other words, the specific type and magnitude of the applied corrections are a measure of how well we match certain experimental details and a guide to further model improvement.

This process of tuning post-shot simulations is difficult, and requires expert judgement and scientific intuition. Post-shot designers have at their disposal many adjustable parameters. These parameters include laser powers at several epochs in the laser pulse, typically 5, and on two independent laser cones. Varying these parameters entails at least a ten-dimensional parameter space to find a better match to data throughout the implosion. The process requires large investments of human time, is often inefficient, and represents a sort of iteratively greedy search for a detailed simulation match to the data.

We choose to take a complementary approach. Rather than invest large amounts of human time guiding the search for imploded plasma conditions that match experimental diagnostics, we can invest computer time to scour parameter space more thoroughly. To do this, we cover the chosen parameter space with thousands of simulations. If this sampling is sufficient, we will find the simulation that most closely matches the experiment, within the variation allowed by the chosen parameters. In addition, we find the family of simulations that forms a neighborhood around the experimental data. Whereas a single post-shot simulation provides a single set of sufficient conditions to match experiment, the ensemble approach provides a notion of sensitivity by capturing the breadth of variations that might also give near matches to experiments.

The ensemble technique provides additional advantages. While our experimental data is very sparse—about 85 cryogenic experiments over 6 years, the ensemble data captures predicted trends over a wide range of parameter space. We capture this variation using a sophisticated fit, or surrogate, that returns an observable value at all points throughout parameter space. This surrogate, then, gives us insight into the sensitivities and topography of observable response as we move through parameter space.

An important example of the utility of this approach arises in the analysis of the achieved thermodynamic conditions in the stagnated assembly. The nTOF diagnostics report a signal width variance that is directly related to the burn-averaged core temperature. Simple analytical models suggest that this burn-averaged temperature cannot exceed approximately 4.2 keV if ignition is not achieved. This limit marks the transition temperature at which alpha-particle energy deposition overcomes radiative loss leading to ignition. On several occasions, the nTOF diagnostics have reported large variances suggesting that this thermodynamic temperature has been exceed in the burning core. Given this important discrepancy, large simulation ensembles were employed to determine the fidelity of the analytical model results.

NIF cryogenic layered implosions have sufficiently high yields to make a comparison of the measured apparent ion temperatures for both DT and DD fusion possible. Naive expectation, and 1D Hydra simulations indicate that the ion temperatures should be the same within the known reactivity temperature variation. The observation is that as the ion temperatures increase, the DT temperature increases more rapidly than the DD temperature.

Simple models indicate that unphysically large velocity variance is required to match observed thermal ion temperature excursion.

The results of a large set of HYDRA-2D simulations including the range of capsule performance congruent with the conditions of the experiments results in features with a DT and DD ion temperature dependence similar to the experimental observations. The inclusion of the experiments' systematic errors shows a statistical agreement between the model and the data (figure 29). This agreement demonstrates that the inclusion of realistic temperature gradients, temperature time dependence and fuel flow velocities can be essential in providing a physical interpretation of the apparently confounding observations.

Zoom In Zoom Out Reset image size

The NIF diagnostics must be able to operate properly during several types of shots with varying neutron yields. Components like CCD cameras and several electronic devices are sensitive to the neutron and gamma backgrounds experienced during shots. Fusion neutrons generated by fuel burn and secondary neutrons resulting from the fusion neutrons interaction with structures present inside and outside the TC contribute to the neutron background. Inside the TC, neutrons interact with the Cryogenic Target Positioner, the warm Target Positioner, and three DIMs. Neutrons also interact with the stainless steel first wall panels, aluminum TC, gunite shield, port covers, and the FOAs. In the meantime, gamma-rays induced by neutron interactions with different structures inside and outside the TC contribute to the gamma background. To simulate the expected backgrounds at different locations, a detailed MCNP [128] model has been developed for the NIF facility. The model includes most of the major components inside the TC and the major diagnostics and structures inside the Target Bay. The detailed Target Bay model is shown in figure 30(a).

Zoom In Zoom Out Reset image size

The Target Bay model has been utilized in the simulation of expected neutron and gamma fluence throughout the Target Bay. The analysis utilized a D–T neutron source spectrum that is expected from a successful ignition target. The neutron source is modeled as a volumetric source filling the hohlraum. The neutron cross-section data were taken from the ENDF/B-VII cross-section data library [129]. Neutron and gamma fluence maps are created for each of the Target Bay floors. A Cartesian mesh size of 30 cm × 30 cm × 30 cm is used to create the maps. To reduce the statistical error associated with the simulation, particle splitting and Russian roulette are used throughout the geometry. The typical statistical error associated with the calculated neutron flux inside the Target Bay was <10%. In addition; radiation damage to electronic devices in terms of dose to silicon has also been calculated for large number of locations inside the Target Bay. These data are being used to provide options for shielding of critical diagnostics and electronics devices to ensure that high precision measurements of neutrons and gammas will be possible up to a 20 MJ yield (7.1 × 10¹⁸ neutrons/shot). During a shot the highest level of radiation is expected near the TC's waist at 7.01 m (23'0'') above ground level. The TCC is accessible through the 5.33 m (17'6'') floor level. Diagnostics and other sensitive components located outside the TC at this floor level experiences the highest level of radiation exposure due to neutrons streaming out of the TC through the large number of diagnostics ports located near the chamber's waist. Figure 30(b) shows a two-dimensional view of the Target Bay's MCNP model at the waist. In this view all the positioners and DIMs are inserted inside the chamber during the shot.

The simulated fluence and dose maps are continually used to predict the background signals as a function of the neutron yield at different locations inside the TB. Figure 31(a) shows the expected neutron fluence values during a 20 MJ shot. The neutron fluence shown in the figure represent contribution from all neutron with energies >100 keV. Contribution from neutrons with energies <100 keV was ignored since, at this low energy, neutron contribution to energy deposition and radiation damage to electronics is minimal. As shown in the figure, very high neutron fluence >10¹³ neutrons cm⁻² is expected inside the TC during a 20 MJ shot. The fluence outside the chamber ports and in line-of-sight with the TCC is ∼10¹² neutrons cm⁻². Fluences at locations shadowed by the chamber's gunite shield are lower by an order of magnitude. It is also important to note that when target positioners and the DIMs are inserted inside the chamber during a shot, they provide self-shielding to components located outside the chamber.

Zoom In Zoom Out Reset image size

To identify the risk to sensitive electronics, another set of maps showing the expected dose to electronics in term of Gy (Si) were also created. Figure 31(b), shows the expected neutron dose values during a shot near the TCC. A high dose of 10⁴–10⁵ Gy is expected near the target (near TCC). The dose drops to 10–20 Gy outside the TC (in the line-of-sight) and to ∼1–2 Gy behind the chamber's gunite shield. Similarly, two more additional sets of maps were created for the simulated gamma fluence and gamma dose throughout the Target Bay. Figures 32 and 33 show the maps generated near the TCC.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

As shown on the figures, neutrons interacting with the gunite generate a large number of gammas outside the chamber. Even though the levels of fluence and dose generated by gamma exposure is somewhat similar to neutron exposure, the impact of each on damage to electronics is quite different. Dose due to neutrons with energy greater than 4 MeV is the major source for expected permanent damage to electronics. However, gamma dose could be an important background contributor to specific diagnostics. Figure 34 shows the calculated dose per unit fluence as a function of neutrons energy and for a 1 MeV gamma source.

Zoom In Zoom Out Reset image size

The simulated fluence and dose maps are used to identify radiation susceptible components based on their known radiation hardness. Such components are tested to a fluence greater than that a component would experience during operation. The testing consisted of component irradiation using a ⁶⁰Co gamma source and accelerator-based irradiation using a 14 MeV neutron source. The Edwards Accelerator Facility located at the Ohio University was used for component testing [130]. Testing of Multiple CCD and CMOS cameras, x-ray generator controllers and digital oscilloscopes indicated that most digital components will suffer from upsets and loss of data at a 14 MeV fluence of 1–5 × 10⁸ neutrons cm⁻² or a dose of 1–5 mGy. The tested components started to fail at a fluence of 1–5 × 10⁹ neutrons cm⁻² or a dose of 10–50 mGy. These results indicated the need for shielding some of the vulnerable components during high yield shots or limiting their use to a lower neutron yield. Predictions for the survivability of tested components are derived from the simulated dose maps and component tests results.

The neutron diagnostics (nTOF, MRS, NIS, NAD) provide the data for the understanding of the overall neutron energy spectrum and the dynamics of the fusion reaction volume in the ICF DT fusion reactions [73]. The nTOF diagnostic fielded at different lines of sights and complementary NI provide key diagnostic information of the symmetry and shape of the fusion burn volume. The data acquired from these instruments allow quantifying the yield and peak shape and burn volume dimensions in comparison with 3D simulations. This also characterizes the NIF as a unique high brightness neutron source for studies of neutron induced nuclear processes and cross section measurements in a HEDP. The understanding of the neutron spectra and consistent simulation of the plasma conditions are key to neutron induced reaction studies. This enhanced understanding in turn will also guide the design of experiments utilizing hot (keV), high density plasmas for non neutron producing nuclear reaction and cross section studies. The evaluation of neutron induced cross section measurements requires a quantitative understanding of the background from room return. First measurements show that the NIF conditions provide for target configurations at TCC in an environment almost free from room return. In the following, key observations from neutron spectrum measurements are discussed.

An ICF DT implosion experiment produces primary neutrons, down-scattered neutrons as well as neutrons from various secondary and tertiary scattering processes within the hot-spot and remaining cold fuel. A typical neutron spectrum is composed of a primary DT peak at 14.1 MeV, a clean down-scattered spectrum between 10 and 12 MeV, a DD neutron peak at 2.45 MeV, and high energy (15–30 MeV) neutrons produced from up-scattered D and T undergoing in flight fusion reactions. The reactions contributing to neutron production are listed in table 1 in the introduction.

The position and shape of the DT and DD neutron production peaks give insight into the temperature and velocity distributions of the hot-spot [131]. Measurements of the tertiary neutron spectrum above 15 MeV, which is on the order of 10⁻⁶ relative to the primary DT fusion neutrons, is challenging due to their low intensity and interference from background γ-rays being produced by 14.1 MeV neutrons along their flight path. The spectrum of those tertiaries provide can provide additional information about areal density and stopping power models, which is crucial for understanding contributions of alpha heating to the overall yield of the shot. Activation materials (Tm, Bi) with thresholds above 15 MeV for (n, xn) reactions, are used to verify the tertiary neutron yield.

The DT and DD fusion neutron peaks are simultaneously observable with the nTOF detectors; the continuous TT neutron spectrum (0–9 MeV) is dominated by the down-scattered neutrons from the primary DT fusion neutrons (figure 35).

Zoom In Zoom Out Reset image size

The nTOF diagnostics measures the peak width of the DT and DD fusion neutron peaks, which relate to temperature and flows in the plasma. It also allows determination of the ratio of down-scattered neutrons relative to the un-scattered flux, which is directly related to the areal density of the cold fuel. Measured shifts in mean peak energies determine the magnitude of bulk motion of the burning fuel [74]. The NAD diagnostics serve to calibrate the neutron yield of the nTOF detectors. Additionally, they provide an activation profile using multiple locations around the TC allowing inferring a combination of ρR distributions and bulk motion. On average, measured yield, ion temperature and down-scatter fraction from the independent MRS, NAD and nTOF diagnostics show agreement [73]. The availability of measurements in several lines-of-sight provides quantitative information for implosion symmetry studies on single shots [132, 133].

The neutron energy in the lab system is obtained from the reaction kinematics as:

$$\begin{eqnarray*}&&{{E}}_{{n}}=({Q}+{K})({{m}}_{{a}}{{c}}^{2}+{Q}/2+{K}/2)/({Q}+{K}+{{m}}_{{\rm{D}}}{{c}}^{2}+{{m}}_{{\rm{T}}}{{c}}^{2})+{{p}}_{{n}}\,{{v}}_{{\rm{cm}}}.\end{eqnarray*}$$

The equation describes the observed energy of a neutron resulting from a single DT fusion reaction. The first term is the neutron energy in the center of mass system (CM) expressed as a function of the kinetic energy K available in the CM system and the resulting Q value of the D + T → α + n reaction and the masses of the particles. The second term accounts for the neutron energy change from CM system into the lab system, with p_n being the neutron momentum vector in the CM system and v_cm the velocity vector of the CM system relative to the lab system; the reaction kinematics is depicted in figure 36. The energy obtained from the relative kinetic energy K is distributed along the Gamow peak of the reaction, with the most likely reaction energy K being ∼5 kT for a 3 keV temperature. The position of the DT peak depends thus on drift velocity and thermal motion of the hot-spot. The center of mass drift velocity of the reaction (burn) volume, the relative kinetic energy of the fusion reactants and the scattering of neutrons in the peak and outside the peak region by the compressed fuel, affect the neutron peak in the spectrum [131].

Zoom In Zoom Out Reset image size

The DT peak is a superposition of 14.1 MeV neutrons from the fusion reaction in the reaction volume in the hot spot and neutrons scattered in the dense DT fuel (see figure 36). Fits to the data from reaction volume include the ρR induced down-scattered neutrons to correct for their effect on the width of the peak. Since the energy, K, is similar for both DD and DT reactions, the neutron energy shifts for DD and DT are likewise similar but the fractional increase is 6 times more for DD; this difference implies that the shift in the energy peak due to temperature is much smaller for DT than it is for DD. The actual width of the observed neutron peak depends thereby on a combination of plasma temperature and flows.

A simple model for the analysis of NIF shot data assumes a single temperature of the burn volume and a single overall motion. In this case, the width of the peak depends only the temperature but the peak position depends on a combination of temperature and motion of the hot spot. A more realistic assumption is that there are multiple burning regions which all have different temperatures and velocities. In this case the superposition of plasma flows and temperatures affect both the width and shift of the neutron peak, which can be described as an observed 'temperature' with kT_app = kT + (m_n + m_α)${{\sigma }_{{v}}}^{2}$ [62] in which ${{\sigma }_{{v}}}^{2}$ represents the variance of the plasma flow velocities, k is the Boltzmann constant and m_n and m_α are the masses of the neutron and alpha particle produced by the reaction.

Since m_α is replaced with m_3He in the DD fusion process, the velocity field broadens the DD peak differently than the DT peak. It is indeed observed that for many cryogenic shots the DT peak has a higher apparent temperature than the DD peak. The presence of a variance in the velocity field indicates that the burning plasma is not stagnant; however, attempting to use the DD and DT apparent temperatures to deduce a thermal kT gives an unrealistically low temperature [134], indicating that the difference in DD and DT temperature cannot be explained by velocity variance alone. There are several other factors contributing to a temperature difference in DD and DT reactions: given a temperature distribution, one can only obtain a neutron yield weighted average temperature. Since the DD and DT reaction rates have different temperature dependence the weighting in the average is different. Also, D and T ions may be at different temperatures and the down-scatter losses are different for DD neutrons compared to DT neutrons.

In addition to the peak broadening due to the variance of the velocity field, the mean of the velocity field causes an energy shift of the neutron peak position. A model using the apparent temperature to calculate the mean energy of the fusion peak according to [135] allows the computation of hot spot plasma (HSP) bulk motion from the nTOF data. Each nTOF detector measures the projection of the velocity vector along its line-of-sight; using the data from at least three detectors a velocity vector can be reconstructed. The nTOF data from the high foot campaign allow to infer HSP (fluid) velocities from Doppler shifts between <30 and 60 km s⁻¹ with half of them >60 km s⁻¹. The HSP velocities inferred from x-ray framing cameras are consistent with the nTOF data. The cause for the HSP (bulk) velocities is not fully understood and it might be a combination of drive asymmetry, ice layer asymmetry, fill tube-jetting, localized oxygen concentration or asymmetric mix/flow of the ice layer fuel into the HSP. The nTOF data are supported by the activation asymmetries measured with the FNADs. Shots with intentional drive asymmetry (P1) showed HSP velocities of the expected magnitude. The FNADs show clear evidence for the strong drive directional dependence of high DT ρRs in agreement with simulations.

A more general description of the neutron spectrum uses the moments of the spectrum and their related cumulants [131] where the first cumulant (mean of the spectrum) relates to the position of the peak due to temperature and HSP bulk motion. The second cumulant, the variance of the spectrum, is designated as the apparent temperature above, which is a yield weighted average of different fluid elements having varying velocities and temperatures. The 3rd and 4th cumulant, skew and kurtosis respectively, give additional insights into the temperature and velocity distributions. This analysis should be done on the neutron momentum spectrum, since a single temperature is most Gaussian like in neutron momentum space, allowing skew and kurtosis to show the deviation from the single temperature case. A theoretical analysis of the neutron spectrum in the two-product thermonuclear (DT, DD) fusion, including effects from a distribution of fluid elements at different temperatures and velocities, is given in [131]. The spectrum shows a down-shift in energy of the centroid energy of the neutron peak for non-igniting NIF-implosions in which the burn occurs mostly during the implosion phase. The analysis furthermore concludes that kurtosis and skew should be universal features of neutron spectral peaks. Measuring the skew and kurtosis of the birth spectrum is complicated by the presence of the down-scattered neutrons and the need to know the IRF extremely well. Current efforts focus on improving the IRF. If skew and kurtosis are measurable, it will allow determining the effect of successive moments of fluid temperatures and velocities on the peak shape in the momentum spectrum. The peak shape features should be distinctively different for the ignition and non-ignition case [131].

A new study of the neutron spectrum from low energy T(T, 2n)α reactions utilizing T-filled fusion targets at the NIF has been reported [136]. The spectrum provides information about interactions among final-state particles and the T(T, 2n)α yield, which is important in ICF experiments studying hydrodynamic mix [137]. The particle spectrum results from three-body decay and is sensitive to different breakup mechanisms. Large uncertainties have remained in the measured spectrum at low reaction energies (E_c.m. < 100 keV), due to a shortage of accelerators with tritium beams, and disparate accounts of its underlying breakup processes exist [138–141].

The spectrum from the T(T, 2n)α reaction has been described previously by statistical and/or sequential processes. Sequential decays proceed through intermediate states formed by n–α (⁵He) or n–n (correlated di-neutron) interactions. For transitions made through the ⁵He state, the neutrons emitted first are spread over the decay width of the intermediate state by energy-momentum conservation. Discussions in the literature of sequential decay through ⁵He have not included interference effects arising from a final state with identical fermions. As reported in [136], the sequential decay process was formulated within the R-matrix framework [142], where the necessary interferences could be incorporated. The modeled spectrum revealed a substantial change arising from anti-symmetrization.

Wong et al [143] determined from their measured neutron spectrum at E_c.m. = 250 keV that 70% of decays went by correlated di-neutron emission, with the remaining strength divided between the ground (20%) and first-excited (10%) states of ⁵He. As stipulated in their paper, the branching ratios were directly tied to an assumption of a strong di-neutron channel.

A previous work [141] has claimed to find no evidence of n–α interactions in the spectrum; the missing strength was explained by an energy dependence in those channels. To address this controversial issue, the ICF type DT implosion of a tritium-filled capsule was combined with the time-of-flight technique to show clearly a peak corresponding to the ground-state channel of ⁵He at low energies.

Using time-of-flight detectors, an yield-weighted ion temperature of k_BT = 3.3(3) keV (or an effective energy of E = 16(1) keV) within the burn volume was extracted from thermal broadening of mono-energetic (14 MeV) neutrons produced by the T(D, n)α reaction. A neutron yield more than 10¹³ was produced over the ∼200 ps duration of the thermonuclear reactions, affording excellent statistics and timing. The neutron spectrum was recorded by two 10 cm × 5 cm xylene scintillators, positioned over 20 m from the TCC along separate lines of sight. Detectors were operated in current mode and their signals were digitized with 1 GHz oscilloscopes. Interpreting current mode data requires knowledge of the impulse response functions (IRF) from detectors. As the light decay time of the scintillator is rather insensitive to particle type, short pulses (100 ps) of x-rays have been used to measure the IRFs. Measurements have also been performed with capsules containing a 3:1 ratio of deuterium-to-tritium. These data serve to establish the neutron-induced background from residual deuterium in the capsules, and show a relatively flat baseline several decades below the preceding T(D, n)α peak (figure 37).

Zoom In Zoom Out Reset image size

In the analysis, the modeled R-matrix spectrum was fitted to data from both detectors simultaneously. This method increases the neutron statistics while averaging over systematic differences between setups. Within the fitting algorithm a trial spectrum was first thermally broadened, then adjusted for source attenuation/scattering and light production. Next, for each detector, the modified spectrum was transformed from energy into time of flight and convolved with an IRF (figure 37 top spectrum) A χ² minimization of the resulting spectrum to data determines free parameters with the MINUIT2 package [144].

Digitizer noise from scopes represented the main source of statistical uncertainty in the data. Fluctuations due to neutron statistics were much smaller, as ≈10⁶ events were recorded within the region of interest. Systematic uncertainties have been investigated for ion temperature, source attenuation/scattering and light production. Separate analyses were undertaken for each systematic error at its 1σ deviation leading to a total error of 16%. Branching values for the 3/2⁻ and 1/2⁻ partial waves determined from the procedure were 29 ± 16% and 71 ± 16%, respectively. Note that the interference between partial waves is found to have little effect on the branching ratio. While differences between the fitted curve and data appear in several locations, the model describes the main features in the spectrum quite well, particularly in the region of the spectrum that is least susceptible to systematic errors, E ≤ 4.5 MeV.

New spectra from ICF experiments with enriched tritium targets have been measured using the NIF. These measurements provide conclusive evidence for n–α interactions in the ³H + ³H neutron spectrum produced by thermonuclear environments, through the first direct observation of a neutron peak from transitions via the ground state of ⁵He. A description of the data was obtained with an R-matrix model that accounts for the coherence between intermediate states and fermion symmetry. No previous group reporting the spectrum has considered these interferences, although as shown in a letter [136] their effects are substantial. Fits of our model to the spectrum have shown a large contribution from the 1/2⁻ partial wave in ⁵He, as opposed to the dominant role assigned to either correlated di-neutron emission [143] or statistical decay. Future work is planned on understanding the neutron spectrum using new di-neutron calculations and/or sequential s-wave emission that take interference effects properly into account. We acknowledge the limitations of this analysis since it represents a two-body approximation to what is strictly a three-body problem, and hope these new data improve evaluations of fusion experiments and encourage further theoretical work.

The ICF environment supplies a unique opportunity to study the particle energy distributions of three-body final states, as the particles may emerge from the low-mass plasma with very little energy loss. In addition, the ICF facilities at NIF and OMEGA have the capability of working with tritium, an initial-state reactant in several interesting and important reactions yielding three-body final states.

For example, the neutron energy spectrum from the T(T, 2n) α reaction has recently been measured at NIF at an effective E_c.m. of 16 keV [136]. The experimental data show a narrow peak at E_n = 8.7 MeV corresponding to the ⁵He ground state and a broad continuum at lower energies resulting from the ⁵He first excited state and other mechanisms such as di-neutron emission.

We have developed an R-matrix model that includes interactions between all pairs of nuclei in the final state [13]. For the T + T case, this implies the nα interaction, including the unbound 3/2⁻ ground and 1/2⁻ first excited state of ⁵He, and the n–n interaction. The calculation also incorporates angular momentum conservation and fermion symmetry. The energy distribution of particles emitted by reactions proceeding to unbound states can be described using R-matrix methods [145], which describes the particle emissions as sequential two-body decays. Our approach also builds on the work of [146–149] and allows different reaction pathways and orders of emission to interfere coherently.

For the T + T case, we have assumed a J = 0⁺ initial state and considered four decay channels: the 1/2⁺ (l = 0) nα channel, the 1/2⁻ (l = 1) nα channel, the 3/2⁻ (l = 1) nα channel, and the l = 0 nn (di-neutron) channel. The interactions in the various two-body channels are well constrained by two-body scattering data. The resulting neutron energy distributions, where each channel is considered individually, are shown in figure 38. The primary (first neutron), secondary (second neutron), exchange (order-of-emission interference), and total contributions are shown. The experimental neutron energy spectrum [136] was then fitted assuming these four channels (and their interferences) were present. As described in [13], an excellent fit to the spectrum can be obtained. Our preferred fit is displayed in figure 39, as a Dalitz plot showing neutron–α energy correlations. The vertical band at E_n = 8.7 MeV and the diagonal band in the lower left part of the ellipse are due to the ⁵He ground state. The concentration of strength at the top of the ellipse, where E_α = 3.8 MeV, is due to the di-neutron.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Our model can in principle describe any three-body final state, provided that the two-body scattering interactions (phase shifts) are known. The feedings of the various reaction channels are treated as free parameters. The reactions ³He(³He, 2p) α and T(³He, np)α, which are related by mirror or isospin symmetry to T(T, 2n)α, are also presently under study at ICF facilities. Using these symmetries, predictions can be generated using the feeding parameters determined from the T(T, 2n) α fitting. Such a prediction for the ³He + ³He proton spectrum is shown in figure 40.

Zoom In Zoom Out Reset image size

In the future, several issues could be clarified by improved experimental data. It would be very useful to extend the T + T neutron spectrum measurements to lower neutron energies to better constrain the fits and to possibly observe the double-humped structured predicted below 2 MeV neutron energy that is associated with the ⁵He ground state (see the 3/2⁻ nα panel in figure 38).

A fully documented measurement of the α-particle spectrum from T + T would also be valuable, particularly if the spectrum could be measured up to the endpoint, where the di-neutron contribution is maximal. It would also be interesting to study the dependence of the spectrum on the energy in the entrance channel, as there is some indication the ⁵He ground state peak is more prominent at higher entrance channel energies [150]. On the theoretical side, it would be interesting and useful to extend our formalism to include the energy dependence in the initial state. We also expect that these methods can be applied to additional reactions or spectra with three-body final states, such as the ¹¹B(p, 3α) reaction.

Experimental studies of charged-particle energy loss have a long history with accelerated ion beams and cold, neutral targets. A combination of Bethe's formula and empirical models are used to describe this experimental dataset [151]. More recently, accelerator-based measurements have been performed with plasma targets generated by discharge tubes and lasers ([152, 153] and references therein), which have covered electron densities of 10¹⁶–10²⁰ cm⁻³ and temperatures (k_BT) of 1–200 eV. Denser plasmas (10²³ cm⁻³) have been investigated with 14.7 MeV protons produced by ³He(d, p)α reactions in ICF experiments [154, 155], either by passing the protons emitted from an implosion through a colder (32 eV), laser produced plasma [154] or by observing the energy downshift due to the hotter (0.5–4 keV), thermonuclear plasma [155]. Yet even denser plasmas (10²⁴–10²⁶ cm⁻³) have been studied with ICF experiments via neutrons from T(D, n)α reactions with energetic (E ≫ k_BT) tritons or deuterons [54, 156]. Several energy loss models exist to treat the fully ionized hydrogen plasmas of ICF, e.g., see [156–160], but the measurements necessary to test their predictions are only now becoming accessible.

Stopping power investigations of the dense plasmas relevant to ICF experiments benefit from indirect approaches to measure energy loss, as the plasma dimensions often exceed the ranges of the charged particles emitted within the fuel. Neutrons avoid the range limitations as well as spectral distortions from remaining ablator and the strong electromagnetic fields that are due to laser–plasma interactions in the regions surrounding the compressed ICF capsule. Moreover, the neutrons serve as a relatively unperturbed diagnostic of the reacting particles' energies.

Indirect drive ICF experiments have used both gas-filled deuterium and cryogenically layered DT targets to study charged-particle stopping in plasmas, under conditions of weak to strong degeneracy (k_BT/E_F = 100–2) and coupling (Γ = 0.001–0.1). The weakly degenerate, weakly coupled plasmas of ICF hotspots can be studied with the gas-filled targets. Thermal D(D, p)T reactions within the hot core formed by imploding a deuterium-gas capsule create an isotropic and nearly monoenergetic source of 1 MeV tritons, and a small fraction (10⁻²) of these tritons initiate T(D, n)α reactions before either thermalizing or exiting the deuterium plasma. The secondary T(D, n)α neutron spectrum observed from an implosion reflects the velocities of interacting tritons in addition to the known neutron angular distribution and two-body kinematics of the reaction which transform these velocities into the spectrum. The stopping power acts on the velocities through their magnitudes, and affects the distribution of reaction energies, as the interaction probability changes with triton energy in proportion to the T(D, n)α cross section and inversely to the stopping power.

Radiation-hydrodynamic (HYDRA) simulations [150] are used to calculate the T (D, n)α neutron spectrum by simulating the time-dependent production and transport of tritons, from their origins in the hot core to their interaction points throughout the plasma. The standard prescription in HYDRA simulations treats charged-particle energy loss using Maynard and Deutsch's [156] version of the random-phase Born approximation for electron stopping and a binary collision description for ions [160]. Another model frequently used to calculate electron stopping power under these conditions, the Fokker–Planck formulation given by Li and Petrasso [157], was included for comparison.

Figure 41 presents the results of the neutron spectral analysis. By further reducing the energies of the tritons, the larger stopping power of Li and Petrasso's model increases the overlap between the energy distribution and the strong resonance in the cross section. The added reactions occur at lower triton energies, populating the neutron spectrum around 13–15 MeV, and makes the area normalized spectra for Li and Petrasso appear slightly narrower than Maynard and Deutsch in figures 41(c) and (d).

Zoom In Zoom Out Reset image size

Order-of-magnitude improvements in the neutron spectroscopy of secondary T(D, n)α reactions have been accomplished with the NIF. This advance now allows investigations of charged-particle energy loss using the neutron spectrum, and extends studies of stopping power to plasmas relevant to ICF, in which an understanding of α-particle heating is necessary to evaluate the performance of experiments and guide future designs. The general agreement found between the measured and simulated neutron spectra provides the first validation of several frequently used stopping power models for these plasmas.

The observed variations in FNAD activities, DSRs, and apparent yields measured along different LOS give information on the underlying capsule areal density distribution at bang-time. This distribution may be extracted from the data for each shot by fitting the nuclear measurements to a model comprised of a spherical shell whose size is determined by the NIS images and whose directionally dependent areal density is specified by the 9 coefficients of a spherical harmonic series summed to L = 2. The fit parameters are the 9 coefficients and the birth yield. The fitting function uses a Monte-Carlo to transport DT fusion neutrons from the hot-spot through the trial areal density distribution. Non-interacting primary neutrons and elastically scattered secondary neutrons are tracked to the detectors and counted; the results are used to predict values for the FNAD activities, the DSRs, and the apparent yield measurements. Nine additional Monte-Carlo runs, each with a single coefficient perturbed slightly from the trial value, are used to compute the derivatives of the predicted measurements with respect to each of the parameters. A modified Levenberg–Marquardt minimizer uses the predicted values and their derivatives to calculate a new set of areal density coefficients and birth yield attempting to minimize ${\chi }^{2}$ (=sum of the squares of the differences between the predicted and measured FNAD activities, DSRs, and apparent yields over the squared errors). The new parameters are fed back to the Monte-Carlo for another step and the process is repeated until ${\chi }^{2}$ ceases to get smaller.

Figure 42 shows the results for N151007-001-999, an HDC layered shot whose FNAD activity distribution as well as the measured DSRs displayed a large north–south asymmetry. The best fit areal density, figure 42(a), indeed varies from 1.5 ± 0.1 gm cm⁻² at the north pole to nearly zero in one spot on the equator. The other panels in figure 42 show the measured data (blue points) as a function of the cosine of the polar angle and the Monte-Carlo predictions (red points) using the best fit areal density. The fit had 35 measurements with 10 parameters and achieved a reduced ${\chi }^{2}$ of 0.9.

Zoom In Zoom Out Reset image size

It is seen that this model fits the data very well showing that the measured activities, DSRs, and yields for this shot are consistent given the appropriate fuel areal density distribution.

This is true for most other layered DT shots which also obtain good fits using this method as seen from the ${\chi }^{2}$ distribution of 26 shots shown in figure 43. The fact that the data fit this simple physical model tells us that the nuclear measurements are providing a consistent picture of the underlying areal density distribution. Current work is focused on how to use this method's best fit areal densities in more sophisticated HYDRA simulations.

Zoom In Zoom Out Reset image size

The CH shell ablator areal density in an ICF capsule can be diagnosed by measuring the γ-rays from inelastic neutron scattering of ¹²C [50, 161]. Experimental detection of the time-integrated γ-ray signal at 4.4 MeV using threshold detection from four gas Cerenkov cells provides a direct measurement of the ¹²C areal density near stagnation (see Herrmann article). Neutron spectrometers can in principle also probe the ablator areal density through measurements of neutrons that have lost energy through scattering in the CH ablator. However, for cryogenic DT shots, the down-scattered neutron signal is entirely dominated by down-scatter from the dense DT fuel layer. For surrogate implosions with the DT ice layer replaced with a thicker CH shell, comparisons can be made between GRH and neutron spectrum ablator areal density measurements, showing good general agreement between the two techniques [162].

GRH data collected from a recent high neutron yield campaign at the NIF reveals two general trends: less remaining ablator mass at stagnation and higher shell density with increasing laser drive. The findings are based on an analysis of experimental results from this recent high-yield campaign and the previous NIF Ignition Campaign (NIC), where Si-doped hydrocarbon ablators were used [163]. The laser drive was modified from the four pulses in the NIC to three with a higher energy first pulse ('high foot') in the high yield campaign. Physical estimates and radiation hydrodynamic simulations suggested that this choice of laser drive would suppress fuel ablator material mixing by reducing the compressibility of the DT fuel. The loss of compressibility with the 'high foot' pulse decreases the likelihood of ignition, but increases the overall stability of the implosion [164]. An increase in neutron yield by a factor of ten and very little ablator mix for the 'high foot' implosions was obtained compared with the best performing implosions during the NIC. For the 'high foot' pulse, the fuel areal density declined by 30%. In addition to the laser drive alteration, three different initial ablator outer radii were investigated: a thicker shell matching that used in the NIC with 195 μm, and thinner shells with 175 and 165 μm radii. The variation in hydrocarbon shell thickness was selected to systematically increase the maximum implosion velocity without further increase in the laser drive power. Since these implosions produced higher neutron yield, there is interest in quantifying the implosion characteristics in detail.

A validated three-dimensional static model used in earlier NIF campaigns was applied to the interpretation of the experimental data [117]. This model correlates the NIF implosion diagnostic information to deduce the thermodynamic properties of the burning core, the spatial density distribution of the DT fuel and the remaining hydrocarbon ablator to generate a self-consistent description of the time-averaged implosion conditions. A three-dimensional volume is derived from neutron images; assuming isobaric conditions at implosion stagnation, the hot core density and temperature variation are obtained; and the directional nuclear scattering and activation data is used to reconstruct the shell and ablator density distribution. The original DT fuel mass is conserved but the remaining hydrocarbon mass is variable.

Of special interest in the work described here is the use of the time-integrated 4.4 MeV ¹²C γ emission to fit the hydrocarbon areal density in the context of this model and thereby derive a better qualitative understanding of the stagnation conditions. Constraints for the ablator areal density arise from estimates of the remaining hydrocarbon ablator mass [165, 166] and the ablator mass mixed into the burning core [167]. In the case of the high-yield series of implosions [164], very little ablator mix was observed in the burning core, implying that the remaining ablator material is either located in a compressed shell around the DT fuel assembly or entrained in the fuel assembly, or both. The ablator mass at stagnation is typically estimated from so-called convergent ablator implosions, which rely upon x-ray back-lighting to measure capsule velocity and remaining mass. These companion implosions used the same laser drive conditions but replaced the DT ice layer with a comparable mass of ablator hydrocarbon. The selected cryogenic implosions varied the laser energy, power, hohlraum lining material and ablator thickness to increase the incoming shell velocity, which induced significant differences in the remaining mass and the GRH-derived ablator areal density.

Data fits described in [50] using the three-dimensional model to six of the high-yield implosions were performed for two of the thicker, two thinner and thinnest capsule configurations and one of the better performing NIC implosion experiments was included for comparison. The original static model was modified to replace the previous constant hydrocarbon ablator region with a quadratic exponentially decaying functional form [50]. As an example, density profiles in a fit are plotted in figure 44. An exponential form is physically reasonable since the hydrocarbon layer is releasing material without a boundary constraint after the shock passage that creates stagnation conditions while the quadratic decrease accounts for the spherical geometric contribution. Furthermore, the use of this simple fitting form provides some initial qualitative insight into the differences among the chosen implosions. The analysis assumes an implosion assembly consisting of the DT gas and DT fuel layer, with the remaining hydrocarbon ablator conformal with the exterior of the DT fuel layer. The DT regions are assumed to be uncontaminated by ablator material and the exterior ablator region is described by a linear exponential form decreasing quadratically beyond a radius r₀ and matched to a Gaussian form that describes the DT fuel density

$$\begin{eqnarray*}&&\begin{array}{l}\rho ({r})={\rho }_{{\rm{DT}}}({{r}}_{0}){({{r}}_{0}/{r})}^{2}\exp [-\alpha ({r}-{{r}}_{0})]\theta ({r}-{{r}}_{0})\\ \,\,=\,{\rho }_{0}\exp [-{b}{({{r}}_{0}-{{r}}_{a})}^{2}]\,{({{r}}_{0}/{r})}^{2}\exp [-\alpha ({r}-{{r}}_{0})]{\rm{\Theta }}({r}-{{r}}_{0}),\end{array}\end{eqnarray*}$$

where r₀ is the matching radius and α is the inverse decay length. The ablator density at r₀ is required to match the compressed fuel Gaussian form which has a peak density of r₀, with a full width at half-maximum b, and radius at peak density of r_a. The Gaussian form is independently determined by fitting the x-ray and nuclear diagnostics, so the three parameters r₀; b and r_a, are known. The remaining two parameters, α and r₀, are determined using the GRH measured hydrocarbon areal density, ρ_CH, and the estimate of the remaining ablator mass, m_CH. The standard 'Heaviside' step function, Θ(s), is zero for s < 0 and one for s ≥ 0. The angle-averaged areal density constraint yields ρ_CH = 1/4π ∫dφ∫dθ sinθ ∫dr ρ(r) = ρ_DT (r₀) r₀(1 + αr₀E_i(−αr₀)) where E_i (x) is the exponent_ial integral function. Application of the remaining mass constraint produces a simple relationship between the matching radius and inverse decay length

$$\begin{eqnarray*}&&{{m}}_{{\rm{CH}}}={\rho }_{{\rm{DT}}}({r}0)\int {\rm{d}}\varphi \int {\rm{d}}\theta \,\sin \,\theta \int {\rm{d}}{r}\,{{{r}}^{2}}_{0}{\rm{\Theta }}({r}-{{r}}_{0})\exp [-\alpha ({r}-{r}0)]=4\pi {{{r}}_{0}}^{2}/\alpha {\rho }_{{\rm{DT}}}({{r}}_{0})\end{eqnarray*}$$

since m_CH is known experimentally. Replacing α in above produces a nonlinear equation for the remaining fit parameter, r₀. This nonlinear equation is readily solved by standard numerical techniques and a comparison of the selected implosion experiments is presented in [50], where the DT fuel layer parameters are assumed to be r_a = 50 μm, b = 10⁶ cm⁻², and ρ₀ = 500 g cm⁻³.

Zoom In Zoom Out Reset image size

Within the context of this over-simplified approach, conspicuous trends may be distinguished as a function of increasing implosion drive. For example, the lower x-ray drive implosions, and lower laser energy drive implosions, have qualitatively similar matching densities and radii, whereas the higher drive implosion shots display much higher matching densities and significantly reduced matching radii. The increased laser drive of these later shots, due to a combination of higher laser energy and the additional uranium hohlraum liner, compressed a smaller amount of ablator mass to an approximate three-fold increase in density at a 10% reduced radius.

There is a clear violation of this trend for the thinnest shell implosion. Since the remaining ablator mass is relatively large but with a small GRH-inferred areal density, the matching density is quite low at the large matching radius and the range, α⁻¹, is large. These values suggest a significantly extended fuel and ablator assembly at stagnation.

In three-dimensional fits to the selected implosions, more realistic conditions are introduced into the model. The density and radius at which the exponential model for the remaining ablator material begins is now a function of radial distance and spherical angle; hydrocarbon material may be introduced into the DT gas and fuel assembly; and other experimental constraints such as the burning core volume and fuel areal density are imposed. Figures 45(a) and (b) [50] show results of the extended fitting procedure by comparing identical cross sectional views of the stagnated assembly in the equatorial plane for the analyzed implosion experiments. The better-performing low-adiabat implosion achieves the most favorable combination of high density mass distribution and relatively uniform assembly symmetry. The subsequent high-adiabat implosions do not attain this uniformly high density distribution. Second, all of the high-adiabat implosions appear to have cold fuel shape distortions and, in some cases, burning core asymmetries. Third, there are two implosions in which the peak hydro carbon ablator layer density is several times larger than the peak densities in the other high-foot implosions. An examination of either the spherical or three-dimensional matching densities or the associated ranges for these shots reveals substantially higher matching densities and much shorter ranges. According to this analysis, the ablator was thus conformally compressed with the DT fuel layer. In all the considered cases, though, the hydrocarbon ablator layer clearly remains outside the regions of highest DT fuel density corroborating the expected behavior of the high-adiabat design in this shot series.

Zoom In Zoom Out Reset image size

A straightforward comparison to the spherical results may be obtained by a simple average of the angular matching densities and radii [50]. Despite the large angular asymmetries in the shell density distributions obtained for many of the implosions, the spherical model captures the qualitative behavior of the compressed hydrocarbon ablator layer. For example, the matching radii are quite similar, except for uneven mass distributions. The trend of increasing matching density with increasing implosion velocity is reproduced in two high-foot series cases but fails for another later thin shell case.

Perhaps indicating the onset of shell disruption, the stagnation characteristics of the very thinnest shell implosion are quite degraded compared to the other high-adiabat implosions but this difference likewise appears in both the spherical and full three-dimensional models. The GRH results are thus in qualitative agreement with the expected experimental implosion design and provide a quantitative measure of the relative difference in stagnation performance. Within the context of these simplified fitting techniques, the GRH-inferred ablator areal densities usefully complement the other stagnation diagnostic signatures and permit an extension of the three-dimensional static model to describe the time-averaged hydrocarbon density distribution at stagnation. In contrast, very little ablator mix into the burning core is detected as measured by the standard x-ray yield diagnostic.

The analysis suggests that further valuable information about ignition-relevant implosions resides in the 4.4 MeV gamma-ray signal. Some of the most important outstanding issues in the approach to achieving ignition concern the amount of DT fuel-ablator mix occurring and the overall compressibility of the assembled DT fuel and remaining ablator. As noted above, these issues might be directly diagnosed using the time-integrated 4.4 MeV gamma-ray emission as captured by the GRH diagnostic.

Finally, we note, the 4.4 MeV GRH data complement the existing solid radiochemical collection (SRC) diagnostic fielded at the NIF. The SRC data measures the ablator hydrogen whereas the 4.4 MeV signal arises from the ablator ¹²C [161].

The study of nuclear reactions at the NIF, requires the collection and analysis of the reaction products, as well as the target matrix (e.g. target material inside a capsule or attached to the hohlraum) during ICF experiments. For solid materials, this is accomplished through the use of catcher foils (solid debris collectors) placed within the TC, as described in section 2.1.9 radiochemical diagnostics (RC). These collectors (typically vanadium, tantalum or graphite) intercept the exploding NIF target assembly during a shot. The target mass, now both atoms and particulates, is collected, removed from the chamber and analyzed using various techniques: photography, microscopy, spectroscopy and spectrometry.

The surfaces of these collectors, orthogonal to the TCC, are exposed to extremely harsh radiation and thermal environments during ICF experiments at NIF. As a result, the exposed area of the collector undergoes significant changes due to melting, impact and ablation processes from the irradiation, along with the dispersal of the target assembly mass [112]. Initial examination of the collector surfaces are typically performed with two nondestructive techniques: optical microscopy and scanning electron microscopy coupled to energy dispersive x-ray spectroscopy (SEM–EDX).

Optical photo-microscopy provides visual information regarding the macroscopic features of the debris distribution and collector surface deformation. For example, shrapnel debris is more commonly observed on the surface of collectors in the polar direction of the NIF chamber. Graphite collectors fielded on this line-of-sight show the cratering impacts and penetration of macroscopic ballistic debris, figure 46(A).

Zoom In Zoom Out Reset image size

For metallic debris collectors, surface features such as melting and movement of surface mass are clearly visible in similar photomicrographs, figure 46(B). Further, splats of melted solid debris, observed even at 5× magnification, are more clearly displayed in the 40× micrograph, figure 46(C).

The superior magnification and imaging of the SEM/EDX coupled spectroscopy system allows more detailed examination of solid debris and the collector surface. The SEM, which produces an image of the surface through raster scanning with a focused beam of electrons, provides detailed surface topography and composition. Magnification of up to 10 000× can be utilized by this technique. The target debris field at NIF is heterogeneous, as evidenced by the diverse debris collected on a single collector, figures 47(A)–(D), which had a collection area of 12.6 cm² at 50 cm from TCC (solid angle of only 4 × 10⁻⁴). Furthermore, extremely valuable information can be extracted from these spectroscopy images, such as evidence of geometric fractionation by two distinct splats of solid debris, figure 47(C). The EDX capability provides an analytical method for qualitatively assessing the composition of collected debris by measuring the emission of characteristic x-rays from regions of interest, figure 47(D). This nondestructive method allows the assessment of the distribution of solid debris, collector surface properties, structural features of the debris and qualitative element identification of the debris composition. The quantitative analysis of solid debris via nondestructive gamma spectrometry is performed routinely for the SRC nuclear diagnostic [113] and has also been utilized in a fractionation study of NIF target debris [168]. The major advantages of this method are that it allows high-resolution, direct counting of the debris collectors with high-purity germanium detectors after minimal sample preparation time. Thus, for an experiment where short-lived radionuclides are of interest to study a nuclear reaction at NIF, this method of prompt, nondestructive analysis is preferred. However, it also has certain disadvantages that make it unsuitable for analysis of many reaction products, as well as all stable/long-lived materials.

Zoom In Zoom Out Reset image size

The reaction products produced at trace levels may be difficult or impossible to quantify with reasonable uncertainties through direct counting of the debris collectors due to: interferences from other reaction products with similar or higher-energy photon energies, and/or more commonly, low production rates for the reaction products of interest. Consequently, radiochemically pure samples will be necessary to increase the sensitivity of nuclear counting measurements by removing unwanted reactions products (including activation products in the collector material). Therefore, destructive methods have been developed for the recovery of solid debris from collectors, entailing subsequent radiochemical separation methods followed by analyses via nuclear counting and/or mass spectrometry.

The surface of the solid debris collectors can be leached using different combinations of mineral acids/bases. Provided that the solid debris is collected within the first few surface μm of the collector [112], the process typically recovers 95+% of the debris. The collection efficiency of the leaching process can be measured through a radioactive component of the collected debris. The leached solution can then be traced with other radionuclides of interest for further processing of the sample. These processes typically require development of new separation methods tailored toward a specific experiment/analyte of interest. For example, a procedure was developed for the recovery of lanthanides from vanadium collectors to address the debris distribution over the VADER collection system [169]. This new method required the separation of microgram amounts of lanthanides from grams of vanadium collector matrix [170]. Similarly, other radiochemical procedures have been developed to address experimental needs, such as the recovery of gold hohlraum debris from graphite collectors [171]. The final radiochemical pure fraction, containing the reaction products and/or stable elements of interest, produces a higher fidelity measurement by addressing the two major disadvantages of nondestructive gamma spectrometry given above. The subsequent instrumental techniques of gamma spectrometry and mass spectrometry (although not limited to these) can produce measurements with detection limits on the order of 10⁵ atoms and 0.1 ppb or better, respectively (assuming normal background levels).

Based on the measurements of ¹⁹⁸Au/¹⁹⁶Au activation ratios collected from activated Au hohlraum material, a new method to determine fuel (DT) and hydrocarbon ablator (CH) areal densities has been developed. The Au ratio combined with the independently measured (nTOF) DSR uniquely determines the areal density of DT and CH in the context of a semi-static 1D capsule model.

The areal density, ρR, of the compressed fuel around the hot spot is typically inferred from the neutron down-scattered ratio DSR = Φ(10–12/13–15) [172]. However when ablator contributions become appreciable, attributable to more remaining ablator mass or entrainment of ablator material into the DT fuel layer, a more accurate method that uses both the DSR and the activated Au ratio for analysis is required. The DSR and Au ratio have different dependencies on ρR(DT) and ρR(CH) due to different cross sections and kinematics (DSR = 0.045ρR(DT) + 0.012ρR(CH)) and (Au ratio = 0.000 74 + 0.0015ρR(DT) + 0.0010ρR(CH) + O((ρR)²) where the ρR values are in g cm⁻². The Au ratio is more sensitive to low energy neutrons since the ¹⁹⁷Au(n, γ) cross section rapidly increases with decreasing energy. The nonlinear behavior of the Au ratio for large ρR values is due to multiple neutron scattering (trapping) in the capsule before the capsule disassembles. The proposed one-dimensional analysis uses the DSR and Au ratio values as constraints, along with several hot spot and cold fuel parameters, to iteratively to fit the unknown fuel and ablator ρR.

Approximately a dozen 'low-foot' ignition shots performed during the NIC in 2011–2013 [173] have been analyzed that show consistently large ρR(CH) values ∼0.7 g cm⁻². A plausible explanation is mixing of the ablator at the interface with the cold fuel. A precise determination of the mix fraction is not possible with this model since the density profiles of fuel and ablator are unknown, as opposed to their measured areal densities.

The radiochemical method for estimating ρR(CH) is complementary to observing neutron induced 4.44 MeV gammas from ¹²C excitation [174]. The radiochemical signature is most sensitive to hydrogen scattering in the ablator whereas the ¹²C gamma ray signal is obviously produced by ¹²C-neutron scattering. Although the methods are complementary, the radiochemical analysis is more flexible since it could be used in cases with alternative ablators, e.g. Be, where no useful gammas are produced [175].

The neutron capture cross section ¹⁹⁷Au(n, γ)¹⁹⁸Au at 14 MeV has been measured on the NIF with high accuracy (10%) and negligible room return by utilizing an exploding pusher platform [176]. This cross section is not well known with only a few separate measurements published [177–181]. The values separate into two clusters of data as shown in figure 48. The evaluated cross section (ENDF/B-VII.1) follows a weighted average of the data in that energy region.

Zoom In Zoom Out Reset image size

An exploding pusher DT gas-filled target with low convergence was used (N130503) to provide a nearly monoenergetic 14 MeV neutron source with small contributions from DD and TT reactions; the measured ρR ∼ 52 mg cm⁻² was small at burn time minimizing neutron down scattering. The time-integrated escaped neutron spectrum was calculated using HYDRA. The static MCNP code with ENDF/B-VII cross sections and the detailed hohlraum geometry gave the expected Au activation ratio. The (n, γ) and (n, 2n) reactions in the Au components were computed by tallying the energy dependent neutron fluence weighted by the Au cross sections. The room-return contribution to the (n, γ) signal was negligible due to the proximity of the Au to the neutron source. The cumulative (n, γ) spectrum shows that 65% of the 14 MeV neutrons account for the capture with smaller but significant contributions from DD, TT and down scattered (Al, Au HR material) neutrons. The ratio of calculated (n, γ) to (n, 2n) reactions integrated over all Au material in the hohlraum yielded an Au ratio of = (1.11 ± 0.05) × 10⁻³, where the estimated error is due to the uncertain spatial origin of the collected Au debris. The measured Au ratio [182] was (9.1 ± 0.6) × 10⁻⁴ and was significantly smaller than the calculated value.

A reasonable procedure for bringing them into agreement is by adjusting the controversial ¹⁹⁷Au (n, γ) value at E_n = 14 MeV, while keeping the ENDF/B-VII.1 cross section values at lower energies. The resulting cross section value σ(n, γ) (E_n = 14 MeV) = (0.95 ± 0.13) × 10⁻³ b (roughly 30% lower than in ENDF/B-VII.1) was used in a modified ENDF/B-VII.1 file for the analysis of layered DT ignition shots [183]. To our knowledge this is the first time an ICF neutron source was used for a neutron cross section measurement.

Nuclear reaction detectors are chemical elements or enriched isotopes that undergo nuclear reactions and may be placed in several locations to provide diagnostic information or basic science data. Isotopes of the noble gases or the halogens (bromine or iodine) are quite versatile as detectors for neutron or charged-particle reactions and the gaseous products are collected with high efficiency in the RAGS system [6] (see 2.1.9).

A significant advantage of gas collection is the high collection efficiency, approximately 95% of the shot-produced gases from the NIF TC is collected in a 15 min sampling period. A second advantage is the ability to provide high precision values for the absolute number of product atoms by combining decay counting on calibrated detectors with isotope dilution mass spectrometry of stable isotope tracers released on each shot. A third is the lack of background contributions for radioactive noble gas species, which simplifies data analysis. To date, we have seen only ¹³³Xe (t½ = 5.24 days) generated from the fission product ¹³³I (t½ = 20.8 h) remaining in the TC after a DT shot containing depleted uranium (DU) in the hohlraum or target capsule. If RAGS collects a sample within 3–4 weeks of a shot that created fission products, a ¹³³Xe background component will be present.

3.2.10.1. Detectors loaded into DT Fuel

3.2.10.2. Detectors intrinsic to or loaded into the capsule ablator

3.2.10.3. Detectors located external to the target assembly

Detectors placed external to the target assembly provide opportunities to measure cross sections, perform basic science experiments and characterize neutron scattering at mounting locations. DU in the hohlraum generates fission products from high energy neutron interactions. RAGS collected the gaseous products and independent fission yields for several xenon radio-isotopes have been determined (see Cassata et al below) [185]. Foils implanted with noble gases, or small rectangular pieces of rigid materials, called postage stamps, may be glued to the outside of the target holder at a fixed and well-known distance with respect to the target center. Neutron reactions on intrinsic or implanted components of these materials generate radioactive species which are released in gaseous or solid debris and collected by RAGS or solid debris catchers (SRC). Gases contained in a re-usable aluminum cell mounted on a DIM may be irradiated and then recovered intact for measurement at the NCF or split for detector calibration. The cell face is located at a radius of 46.5 cm from the NIF TCC. A cell can hold up to 42 cc STP gas, and each DIM can hold four cells. The cells were commissioned in late 2015 using three different xenon isotope mixes: atmospheric xenon; pure ¹³⁴Xe; and a 4:1 mix of ¹²⁴Xe:¹³⁶Xe. Figure 49 shows a gamma spectrum from the third mix.

Zoom In Zoom Out Reset image size

The RAGS system (see 2.1.9 [6]) is a diagnostic apparatus designed to collect, purify, and analyze gaseous reaction products. The system is very efficient at processing TC gases (>95% pumped through RAGS) and is equipped with on-site counting capabilities that enable measurements of nuclides with short half-lives (as short as 10 s). On shots wherein the target hohlraum is lined with DU, D–T fusion neutrons undergo fission reactions with the hohlraum producing xenon (Xe) and krypton (Kr) isotopes and their precursors (Sn, Sb, Te, I, Ge, As, Se, and Br), which subsequently undergo decay to isotope of Xe and Kr in the TC. Here we provide an overview of the gas collection, purification, and measurement procedure used to determine fission yields of Xe and Kr isotopes produced in the NIF TC.

The RAGS diagnostic is equipped with several scintillation detectors for measuring gamma ray emissions from gases trapped at different stages of the system (see 2.1.9). There are two lanthanum bromide detectors on the filter cart, one of which is located adjacent to the exhaust pipe from the water vapor trap, and one of which faces the stainless steel housing surrounding the first hot getter. There are also two high purity germanium (HPGe) detectors, one of which is located above the cryogenically cooled Cu collection foam where Xe is typically trapped, and one of which is flush against the upper wall of the shielded abort tank. The HPGe detector located above the collection foam is affixed to a translatable stage, such that its position can be optimized to reduce detector deadtime. All fission product yield measurements are conducted using the HPGe detectors above the abort tank and collection foam, with data acquisition performed in list mode (i.e., each registered event has an associated timestamp). In [6] a detailed description of the individual hardware components and their operation is given.

Figure 50 illustrates typical gamma ray spectra obtained using the HPGe detectors located above the collection foam and abort tank. Photo-peaks observed in the collection foam HPGe detector spectrum are associated with isotopes of Xe and its daughter products Cs and Ba, with the exception of the 511 keV photo-peak from ¹⁸F decay (t_1/2 = 109.77 min) and photo-peaks from Kr decays 'in flight' as it is pumped to the abort tank. Photo-peaks observed in the abort tank HPGe detector spectrum are associated with isotopes of Kr and its daughters Rb and Sr, except for the 440 keV photo-peak from ²³Ne decay (t_1/2 = 37.2 s).

Zoom In Zoom Out Reset image size

For each nuclide identified, background subtracted time-resolved photo-peak intensities are calculated from the list mode data. Nuclides first appear at the collector cart approximately 20 s after shot time. Decay-corrected photo-peak intensities increase throughout the 15 min collection interval as the fission products are pumped from the TC to the RAGS system in the basement of NIF, and due to ingrowth of Xe and Kr from precursor nuclides that decay in the TC [185]. Because of this dynamic pumping process, the decay-corrected activity ratio of any two nuclides initially reflects the ratio of their independent yields, and generally subsequently approaches the ratio of their cumulative yields. In [185] a detailed description of extracting independent and cumulative fission yields from gas collections at the NIF is provided.

The nearly instantaneous, high-flux neutron source provided by the NIF, coupled with the rapid, on-site gas collection and counting capabilities, enables quantification of nuclides with half-lives as short as 10's of seconds. Thus far, ¹⁴⁰Xe, ¹³⁹Xe, ¹³⁸Xe, ¹³⁷Xe, ¹³⁵Xe, ^135mXe, ¹³³Xe, ^133mXe, ⁹¹Kr, ⁹⁰Kr, ⁸⁹Kr, ⁸⁸Kr, ⁸⁷Kr, and ⁸⁵Kr have been measured. A summary of Xe fission yield determinations is given in [185], and calculations of Kr fission yields are in progress. The experiments conducted to date demonstrate the feasibility of fission product measurements at NIF, and efforts are underway to improve the operational protocol of the RAGS system. Future work using the RAGS system will focus on measurements of charged particle reaction products as implosion performance diagnostics.

The current nToF spectrometer system is optimized for making precise, accurate measurements for neutron kinetic energies 2 MeV ≤ E_n ≤ 15 MeV. Neutrons with energy below and above this region offer the possibility of additional diagnostics for NIF implosions as well as opportunities for measuring nuclear cross sections.

A pair of low (<2 MeV) and high (>15 MeV) neutron energy detectors located at roughly 100 m from TCC would greatly enhance the quality of the neutron spectroscopic data currently obtained in NIF implosions (figure 54). The pair would consist of complementary detectors: a LENS to measure neutrons from 10 eV up to 100 keV, and a high energy NTOF (HENTOF) detector to measure up to 32 MeV. LENS extends the lower range of NTOF spectroscopy down to ∼eV scale. With such a spectroscopic tool, cold fuel temperature, confinement time, and low energy cross sections are measurable. The HENTOF detector will improve the resolution and precision of traditional neutron spectroscopy measurements (DD, TT, and DT Tion, DSR, yield), while providing additional high energy information. The HENTOF can improve measurements of the reactions in flight (RIF) neutron spectrum at much higher energy resolution than possible with existing NTOF spectrometers fielded at the NIF

Neutrons in the energy region 10 eV ≤ E_n ≤ 100 keV are sensitive to the cold fuel temperature, plasma confinement time and low energy cross sections [191]. For ignition capsule implosions, the LENS is determined by the 'leakage' of multiply-scattered neutrons confined in the fuel by the relatively large cold fuel ρR. These neutrons have a spectral shape determined by the cold fuel thermal temperature and a magnitude that is sensitive to the confinement time. Evidence for these neutrons has been seen in the SRC where the ratio of Au isotopes N(^198m+gAu)/N(^196gAu) is collected from hohlraum debris. The capture reaction is dominated by the low energy neutron component whereas the high energy fusion neutrons control the (n, 2n) production. Thus, this ratio provides a yield normalized measure of the spatially averaged cold fuel ρR [113]. The interpretation of this radiochemistry diagnostic currently depends on a realistic model of the low energy neutron flux and confinement conditions.

Measuring the low energy neutrons directly requires transport of those neutrons with high transmissivity while rejecting the photon background along the line-of-sight. The strategy of the LENS diagnostic is to use a highly-segmented detector capable of measuring the neutrons but blind to photons. The particle rate cannot exceed the bandwidth capabilities of the signal digitizers for recording single particles. A LENS comprised of 2 × 2 × 5 mm pixels of 10%-borated plastic scintillator would have high detection efficiency for neutrons in this energy range while being essentially blind to gamma rays [192]. The maximum neutron rate in a pixel is less than 1 event per 10 ns, suitable for single neutron counting. Providing a rough vacuum (about 10⁻⁴ Torr) along the line-of-sight from the TC port to the LENS will also be required.

Neutron energies in the range 15 MeV ≤ E_n ≤ 30 MeV are produced by RIF that provide information on ion energy loss and stopping power in the imploded capsule [193]. The HENTOF spectrometer consists of a large organic scintillator (e.g. bibenzyl which is used in the current 20 m NTOF spectrometers [194]) positioned 20 m downstream of the LENS. A 'cleanup' collimator reduces the background radiations produced by interactions in the LENS detector system. The improved energy resolution of the HENTOF over the 20 m NTOF system, 0.7% from 3.3%, isolates the RIF signal from the primary 14 MeV fusion neutron peak, reducing the background from these primary neutrons. The HENTOF will also provide a more accurate measurement of the primary neutron spectrum down to 1 MeV thus including the important DD-n peak at 2.45 MeV whose physical interpretation is currently challenging in DT-filled capsule implosions.

Inspection of simulated neutron spectra (figure 52) makes it clear that additional implosion physics can be measured with the appropriate new NTOF tools LENS and HENTOF. The left-hand figure shows that RIF neutrons populate the portion of the spectrum above about 15 MeV to roughly 30 MeV. The RIF measurement facilitates inference (coupled with simulations) about charged particle stopping power in partially degenerate plasmas (the cold fuel layer) and hence alpha heating. The right figure shows simulations that illustrate the neutron thermal peak (at around 10⁻⁴ MeV), plotted for different simulated confinement times. Notice that the effective temperature (peak location which is related to the cold fuel temperature) does not change depending on confinement time, but the magnitude changes appreciably.

Zoom In Zoom Out Reset image size

The present diagnostic challenges due to backgrounds that prevent measurement of RIFs and LENS can be greatly reduced by placing two unique (low and high energy) NTOF detectors at 100 m (figure 53). The left-hand figure shows that the late time signal in an NTOF detector is dominated by a nearly constant-in-time gamma ray background. This contribution arises since the detector is too large to effectively suppress gamma ray detection. Thus, the new LENS system should have small volume individual detectors or pixels to suppress gammas while effectively detecting low energy neutrons. As an example, 2 × 2 × 5 mm pixels of 10%-borated plastic scintillator would have a high detection efficiency for thermal neutrons while being essentially blind to gamma rays. Figure 53 (right) demonstrates that the portion of the NTOF signal in the RIF energy (time) region is contaminated by background gamma rays from 14 MeV neutron interactions upstream in the line-of-sight; the gamma ray time-of-flight is less than the nToF to the detector. This portion of the spectrum needs to be reduced by 10⁻⁵–10⁻⁶ (relative to the primary DT peak) to measure the RIF spectrum to ∼10% statistical relative uncertainty. By placing the detector far away from scattering sources at 20 m, the leading edge of the time-of-flight signal has substantially less background. A carefully designed beamline has been shown empirically to be clean down to the ∼10⁻⁵ level. Placing the detector farther away from the TCC would reduce the background to levels meeting the measurement requirements.

Zoom In Zoom Out Reset image size

A high resolution, full spectrum NTOF spectrometer (figure 54) consists of two separate detectors at roughly 100 m from TCC. The beamline will consist of three upstream collimators: one at the TC, one at the TB wall, and one at the building wall. A rough vacuum line-of-sight tank will be required to minimize low energy neutron capture/attenuation in the roughly 70 m of air (air attenuation is only a minor concern for high energy neutrons). The first detector is a LENS situated 100 m from TCC on the equatorial plane of the NIF. The detector consists of highly segmented or crossed fiber boron-loaded scintillator arrays. The scintillators are readout by segmented PMTs and VME digitizers, and are expected to operate in pulse event-by-event mode. The HENTOF detector will consist of a roughly 50 cm diameter slab (or segments) of bibenzyl (like the ones currently used in the 20 m spectrometers [195]) or equivalent organic scintillator that is viewed by 8 PMTs producing excellent sensitivity down to 1 MeV, covering several orders of magnitude in neutron yield. This detector should be situated >20 m downstream of the LENS. Ideally there will be a small collimator between the LENS and the HENTOF to eliminate some of the scattered neutrons/gammas from neutron interactions on the LENS. The HENTOF will operate in current mode, in much the same way as the existing 20 m spectrometers at NIF do.

Zoom In Zoom Out Reset image size

A LENS uses boron-loaded scintillators to detect low energy neutrons with high sensitivity. As an example, consider figure 55 which is a simulation for a 'High Foot' shot at 10¹⁶ total yield. The figure depicts the number of neutrons incident upon the detector in each time bin; notice that the number of neutrons in the 10–100 eV window is substantial (10³–10⁷) depending on whether it was a low or high ρR shot. With a 1%–10% detection probability in these windows, the signal is measurable. Further, the high segmentation allows for a higher detection rate than normal and thus mitigates most of the possible pileup. If small pixels are used, 2 × 2 × 5 mm, the average 1 MeV photon makes only 12 keV electron-equivalent light, while the 4 MeV alpha from the ¹⁰B(n, α) reaction produces 200 keV electron-equivalent light. Thus, a threshold can be placed on the pulse train excluding virtually all the gamma ray hits in the detectors. Another background suppression feature is a 'beam dump' located far downstream, which will substantially reduce the signal backgrounds at late time; however, the requirements on such a 'beam dump' can be relaxed substantially due to the insensitivity of the detector to gamma rays.

Zoom In Zoom Out Reset image size

Utilizing the HENTOF, the RIF neutron spectrum measured at 120 m is cleaner and provides 500% better resolution than measurements at 20 m. The high energy neutrons are produced by two mechanisms that are most interesting: alpha knock-on neutrons (AKN) (neutrons from alpha-up-scattered d, t) and RIF neutrons (neutrons from 14 MeV neutron-upscattered d, t). Of special interest are the AKN which may be a sensitive indicator of alpha heating in the cold fuel shell, a necessary condition for achieving ignition. Figure 56 (left) shows that DT secondary neutrons only extend to about 18 MeV for temperatures of interest to NIF. Beyond that, the spectrum is clean and is dominated by AKN (up to about 21 MeV) and RIF (up to about 32 MeV) [196]. For neutrons in this energy region, a 20 m NIF neutron spectrometer has an energy resolution of about 3.3% and therefore compromises the spectrum measurement (figure 56 (right)). However, a similar detector operating at ∼100 m would have a resolution of 0.7%, yielding better measurement of the neutron energy spectrum details in that region. The higher energy resolution would allow better separation of the AKN compared to the RIF neutrons.

Zoom In Zoom Out Reset image size

Note that the gamma background in the AKN/RIF window is dramatically reduced by moving the detector far from the scattering sources.

In addition to adding a measurement capability above 15 MeV, the HENTOF detector will measure the standard ICF parameters with much higher accuracy. It should be noted that the core of the IRF will be comparably narrow (∼4 ns), so that the DT T_ion measurement becomes more accurate since the width of the peak is now so much larger than the width of the IRF (intrinsic resolution).

A Compton gamma spectrometer, named as the GEMS has been designed to enable the measurement of the prompt γ-ray energy spectrum during high yield DT implosions at the NIF [96, 197]. One example of the prompt γ-rays includes the 16.75 MeV γ-ray line from T(D, γ)⁵He, which provides a direct signature of the total DT fusion yield. Another example is the 4.44 MeV γ-ray line from ¹²C (n, n'γ)¹²C, which is created when energetic DT fusion neutrons inelastically scatter on their way through the plastic capsule ablator. The intensity of the 4.44 MeV is thus directly related to the areal density of the capsule [95]. The prompt γ-ray energy spectrum will provide 'burn-averaged' observables, such as the total DT fusion yield and the total areal density (ρR). The burn-averaged observables from the γ-ray spectrum are unique because these observables are essentially averaged over 4π about the implosion and do not depend on the line-of-sight to measurement devices. Therefore, the GEMS observables will provide a global reference for the line-of-sight-specific measurements, typical of x-ray and neutron diagnostics. The schematic of the GEMS shown in figure 57, consists of a Compton converter (e.g., beryllium), electromagnet, and an electron detector array placed at the focal plane. Incident γ-rays from the implosion target scatter electrons out of the converter plate primarily through Compton scattering [198, 199]. The Compton electrons enter an electromagnet where energy selection takes place. The dispersed electrons enter Cherenkov radiators (e.g., fused silica) where their binned energy is converted to UV/visible photons for detection by photomultipliers (PMT). The GEMS will require ∼5 × 10¹⁴ minimum DT neutrons to measure 4.44 MeV ¹²C γ-rays at 200 mg cm⁻² areal density and ∼2 × 10¹⁵ minimum DT neutrons to measure 16.75 MeV DT γ-rays.

Zoom In Zoom Out Reset image size

X-ray and neutron self-emission images provide diagnostic information useful for determining the hot, reacting volume in ICF experiments. Complementary to these images, gamma ray and scattered neutron images are useful for diagnosing the size and shape of the non-reacting material, which provides insight into the symmetry and integrity of the shell compressing the hot core. Figure 58 illustrates these concepts. Starting with the innermost volume, free-streaming 14 MeV neutrons may be aperture resolved and converted to light in a scintillator which is then recorded by a time gated digital camera, resulting in a projected image of the region where fusion reactions are occurring. If the neutrons elastically scatter to lower energies, which is occurs predominantly in the dense DT shell surrounding the reacting volume, a second camera with a delayed time gate may be used to record the projected image where neutrons are scattering. Finally, if the neutrons undergo inelastic collisions, which excite nuclear levels in a nucleus, e.g. ¹²C(n, n')¹²C*, where the ¹²C* promptly decays by emission of an MeV scale gamma ray, then a third camera system may be used with an advanced time gate that produces an image of the outer ablative shell tamping the DT assembly.

Zoom In Zoom Out Reset image size

Neutron and x-ray imaging have been discussed previously in this article, but it should be recalled that neutrons scatter in the 'hot spot' and the cold fuel regions, as well as the lower areal density carbon ablator. The image brightness (signal) in each element will be reduced at larger radii due to the decreasing flux, and in the case of the carbon ablator, at lower areal density. Simulated gamma images of the carbon ablator show that these images extend to a much larger radius than that of the cold fuel and provide diagnostic information on the measure of compression of the unablated shell.

The size and shape of the remaining ablator material during burn is important to ignition experiment design. If too much ablator mass remains, the implosion will evolve more slowly and generate less internal energy than desired. If too little ablator mass remains, instabilities will result, potentially tearing apart the tamper and compromising the nuclear performance. The ability to assess the size, shape, and remaining mass of the ablator is therefore useful when diagnosing hydrodynamic performance of the experiment during burn.

As mentioned, imaging the remaining carbon ablator material is possible via the inelastic scattering of neutrons on carbon, especially ¹²C(n, n')¹²C*, which will populate the first excited state of the ¹²C nucleus as depicted in the level scheme of figure 59(a). As shown by the blue arrows in figure 59(a), the transition to the ground state results in a 4.44 MeV gamma ray. These gamma rays are emitted virtually isotropic, as shown by figure 59(b) and can be imaged through apertures designed for neutrons. Evaluated nuclear data estimates for the cross section at 14 MeV for this reaction are approximately 200 mb. Simulation and experimental estimates of the remaining ablator during burn are about 200 mg cm⁻², resulting in approximately one 4.44 MeV gamma for every 500 neutrons passing through the ablator. Although this signal rate is slightly less than the observed rate in down-scattered neutron images, it is necessary to use signal multiplexing apertures and a more sensitive radiation converter to obtain the desired gamma-ray image.

Zoom In Zoom Out Reset image size

Researchers at Los Alamos National Laboratory have studied the feasibility of imaging 4.44 MeV photons and have developed a conceptual design for a gamma imaging system at the NIF [200, 201]. In addition to the physics requirements, the principal constraints were to minimize the cost and the impact to the existing NIF imaging diagnostic suite. This work concluded that the penumbral apertures within the NIF NI system were sufficient to provide images with peak signal-to-noise estimates larger than 20 for shots with neutron yields greater than 10¹⁵.

A gamma imaging proof of concept test was performed on the NIF shot N150518. The NI system gates were configured to collect images of primary 14 MeV neutrons and gamma rays interacting in the radiation converter. Figure 60(a) shows the raw image of 14 MeV neutrons through the NIF NI aperture [92]. The image clearly shows the 20 pinhole images and 3 penumbral images. Since the penumbral apertures are large bores with a diameter larger than the source, the resolved neutron fluence is coded into the penumbral portion of the image, requiring a computed tomographic analysis [202] to reconstruct the source. Figure 56(b) shows the image of high energy photons interacting in the converter. Since the converter does not have energy resolution and the camera gate integrates all energies of photons, studies of backgrounds, such as LPI hard x-rays, are required to fully interpret the observed signal. Since the mean free path of the MeV scale photons in the gold aperture assembly is about 1.5 cm, the effective diameter of the neutron pinholes is less than 1 μm, providing a solid angle that is too small to produce useful gamma images, which is confirmed in the figure. Further, since the penumbral apertures have a field of view of about 250 μm, and the aperture is collecting neutron images, the source of the radiation observed in the penumbral images of figure 60(b) must be either from the compressed target, or from inelastic neutron interactions along the aperture line-of-sight and within 1–2 m of the target. This latter source was ruled out in a subsequent experiment using an IDEP, which contains all the same line-of-sight materials as shot N150518, but did not have the compressed carbon present to produce 4.44 MeV gammas. This experiment confirmed that the signal collected in figure 56(b) was from high energy gammas being produced in the compressed target assembly.

Zoom In Zoom Out Reset image size

The conceptual design of a gamma imager on NIF is shown in figure 61. Radiation from the implosion passes left to right through two radiation-to-light converters, with the left most being a coherent array of organic (BCF-99-55) scintillating fibers to image neutrons, and the right most, a coherent array of LYSO crystals to image gammas. The flight path length to these converters is approximately 28 m, resulting in arrival times of roughly 90 and 540 ns for the gammas and neutrons produced at bang time, respectively. This allows fast, O (10 ns) MCP image intensified gated camera systems to image the light from the corresponding converters. Since scintillation light travels both left and right in the figure, two cameras may be employed to add a framing camera capability to the imaging system.

Zoom In Zoom Out Reset image size

Starting in 2018, the NIF will field a dedicated gamma ray imaging system, comparable to that shown in figure 61. The goal of the system will be a measurement of the thickness and uniformity of the carbon ablator that remains at peak neutron production time. The shell uniformity of density and thickness is important to the success of obtaining ignition on the NIF, and an excellent metric for validating current predictive tools. Considerable work has been done to demonstrate the viability of a gamma imaging system with prototype detector systems already developed and tested at HED facilities, such as Omega, and at the High Intensity Gamma Source at the Duke University FEL [203]. The necessary image reconstruction algorithms are available to invert the coded aperture source [204–206].

It has been demonstrated that a HEDP environment with high neutron fluencies at keV temperature and high density conditions in the short pulsed (ps) time domain is routinely achievable at the NIF (see section 2). There is increasing interest in developing a scientific program to utilize these conditions for nuclear science and more specifically for nuclear astrophysics research at the NIF. Several start up projects and research opportunities have been identified in international workshops (e.g. see [207] and are summarized in an NNSA White Paper [208]). A proposed experiment and precursory effort are focused on the measurement of the ¹⁰B (p, α) ⁷Be nuclear reaction rate in the HEDP environment to better than 20% accuracy. The study will be primarily based on the measurement of the characteristic activity of ⁷Be that needs to be collected with high efficiency by collectors positioned near the capsule. This effort is complementary to ongoing accelerator experiments at the University of Notre Dame. These experiments provide accurate benchmark data that are essential for astro-physical cross section networks and theory; the ¹⁰B (p, α) ⁷Be reaction study is also of interest to aneutronic fusion reaction studies. The HEDP environmental conditions in a NIF shot, are determined by ICF oriented diagnostic data for DT fusion and corresponding modeling assumptions. The NIF based reaction rate data can in turn, be utilized for mapping and understanding of the shock dynamics that determines the thermal energy distribution (e.g. Maxwellian). The experiments new tool for innovative nuclear science experiments will open the pathway for lower cross section reactions such as the important ³He (α, γ) ⁷Be reaction in nuclear burning in our sun [209] that defines the solar neutrino flux.

The NIF produced plasma mimics the conditions of the stellar interior, yet the time-scale is dramatically different. Measuring and interpreting nuclear reactions of nuclear astrophysics relevance constitute a fundamental challenge [210]. It is therefore essential to establish benchmark reactions, for monitoring the development of temperature, density and other parameters in the NIF environment. This enables the conversion of reaction data to stellar cross sections and rates, for nucleosynthesis simulations and for scaling of future nuclear reaction measurements. These experiments would also be the first opportunity to test theoretical assumptions about stellar plasma screening which so far only rely on accelerator experiments on low temperature, low density gases.

Nuclear reactions in stars are dominated by charged particle interactions at very low energies that are typical for quiescent stellar burning conditions. The cross sections corresponding to the stellar temperatures are too low to be reliably measured at present accelerator facilities and rely strongly on the extrapolation of existing higher-energy data towards the stellar energy range [211]. It was demonstrated in very low energy experiments at 'background free' underground laboratories, that atomic screening effects cause an enhancement of the reaction yield [212]. On the other hand, reactions between completely ionized nuclei in stellar plasmas are subject to plasma screening through the free electron cloud that will modify the reaction cross section compared to accelerator laboratory data. Screening affects the time scale, the energy production and the nucleosynthesis associated with the specific reaction in question. The standard treatment of the screening conditions relies on the theoretical Debye–Hückel formalism that has not been tested experimentally [213]. The ¹⁰B (p, α) ⁷Be reaction provides a perfect case for testing screening predictions. Initial accelerator work has been done at low energies [214]. The cross section is relatively large, due to the contribution of a resonance at 10 keV, and the possible impact of screening has been evaluated. Complementary to this effort, indirect measurements have been done using the so-called Trojan horse method to probe the nuclear reaction cross section on bare ¹⁰B nuclei [215]. From the experimental point of view the reaction product ⁷Be has a considerable advantage since it does not occur naturally in nature, and it can be relatively easily collected and detected as a shot product.

The measurements of charged particles at the NIF require loading the target capsule or capsule shell material with the needed target material such as hydrogen and heavier mass material. Similar platforms as described in section 2 are then used for the reaction rate measurements.

For the ¹⁰B (p, α)⁷Be experiments the ablation/compression scheme, which heats the target material inside the capsule, is achieved in a direct polar drive configuration (less compression) or in a hohlraum utilizing the 300 eV ablation plasma (more compression). The experiments require reaching a Maxwellian averaged temperature of about 20 keV. For some precursor experiments a very limited DD neutron production can be used as a thermometer. The approaches to achieve stellar burn or ICF are similar except that for stellar burn reaction studies [216] higher number densities of reacting target atoms can be used (e.g., 10¹⁹ atoms of target material in the capsule).

Thermal temperatures of about >20 keV can be reached in a special designed NIF target capsule containing sufficient mass of the reaction constituents of ionized ¹⁰B and H in a plasma. The proposed ¹⁰B/H plasma target would provide a high-density target with a number density of about >10¹⁸ bare ¹⁰B and protons, where about 10¹⁵ could reach temperatures of ≤40 keV. The target material could either be doped in the shell material or it could be the gas load in a capsule with specified shell material.

The first phase of the approach is an evaluation and optimization of the target configuration using advanced hydro-dynamic simulations. As an example, the following configurations will be evaluated: an exploding pusher configuration could be used, where ablation of a low Z-outer shell drives a high-Z inner shell, containing the ¹⁰B/H, inward and thereby moderately increasing the areal density and raising the temperature. The maximum temperature is reached at the point of maximum shock compression as the spherical target converges. The lasers resemble a configuration close to a spherical symmetric direct drive (no hohlraum) with lasers impinging on the target directly from two hemispheres; a configuration utilizing a hohlraum will be simulated as well. Practically the high-Z metal inner-shell shell (e.g. Au) contains ¹⁰B₂ H₆ (di-borane) and the shell is coated with a lower-Z outer-shell (CH plastic) as pusher; this should heat the ¹⁰B/H plasma target to high temperatures (in hohlraum or direct drive configuration).

Simulations for the target capsule scheme and drive configurations will be performed to optimize a possible laser drive and target configuration in terms of the achievable temperature and target density.

An estimate of the reaction rate for the ¹⁰B(p, a)⁷Be reaction using a 1D hot spot model gives yields in the range of Y ∼ 10⁴–10⁵ (assumptions: number densities N of 10²⁰ cm⁻³, reactivity 〈σv〉 of 10–20 cm³ s⁻¹ (T_ion ∼ 30 keV), hot spot volume V of 10⁻⁶ cm³, burn width Δt of 10⁻¹⁰ s); the reaction yield is given as Y_C = N_a N_b 〈σv〉 V_HS Δt, (figure 62). (Note: the reaction rate in the thermalized plasma (high areal density) is R (a + b, NIF) = 〈σ_ab v〉 A_V ρ_a ρ_b V t(s); the rate at an accelerator for a specific beam energy is R (a + b, accelerator) = I_a σ_ab N_b Δx.)

Zoom In Zoom Out Reset image size

The measurement of the reaction rate for ¹⁰B (p, α) ⁷Be as a benchmark requires the development of an efficient ⁷Be collection capability which will enable the reaction rate measurement with good accuracy for the very low (<nano-barn) cross section reaction ³He (⁴He, γ)⁷Be from the solar H-burn (p–p) reaction scheme for further evaluation of the standard solar model.

The proposed scenario for the collection of ⁷Be from the ¹⁰B (p, α) ⁷Be and later the ³He (α, γ)⁷Be reaction is briefly described (the decay of the 3/2-state in ⁷Be is depicted in figure 63).

Zoom In Zoom Out Reset image size

As an example for precursor experiments radioactive (4 micro Curie) ⁸Be is mounted at the hohlraum in DT ICF shots to determine the collection efficiency in ride-along shots. A collection calibration method for the collection efficiency is developed utilizing targets, loaded with Au/Xe and/or DD with known (n, γ) cross sections in a shot prior to the shot with the none neutron producing ¹⁰B/H-target. This baseline experiment allows the reaction rate to be measured relative to the well-known DD reaction cross section.

Other possibilities for the ⁷Be collection could include a 'gas catcher system' with collection of the ⁷Be in an increased Xe background gas in the NIF chamber (e.g., a small exploding cry-coated xenon shell in the target vicinity or filling the whole NIF TC with low pressure xenon to stop and catch the reaction products). We can use the established radiochemical gas analyzing system (RAGS) following a shot. These possibilities are under consideration including an efficiency evaluation.

Development of a high efficiency collection system for non-noble gas reaction products from the exploding target capsule for NIF is a focus of the radiochemical diagnostic at the NIF. The development will enable various other programmatic experiments at the NIF. The development of the overall experimental capabilities (e.g. diagnostic, target, laser conditions) will support a variety of nuclear astrophysics reaction studies, but also demonstrate a target capability useful for stockpile-steward ship programs as well as applied physics programs (e.g., intense neutron source developments involving high density, high temperature non-ignition targets).

Complementary to the reaction rate measurements at the NIF, detailed low energy measurements at the Notre Dame NSL are being performed. Comparison of the rate data allows quantification of the screening effects for the low energy cross sections of the selected benchmarking reactions. Figure 64 shows a recent plot of experimental and theoretical cross sections for ¹⁰B(p, α)⁷B; new accelerator based data and comparison to theory are in progress at ND. The facility is optimally equipped for low energy nuclear reaction measurements with stable and light radioactive particle beams [217]. These experimental conditions are optimized for performing the proposed reaction studies using complementary in-beam gamma ray spectroscopy methods with the high-resolution Germanium-detectors as well as off-line activity. The experimental data are analyzed and fit with the multi-channel R-matrix code AZURE [218] that was recently developed by the Notre Dame group and has been used successfully for extrapolating low energy reaction cross sections with high precision by parallel fitting of multiple reaction channels. This combination of low energy measurement and R-matrix extrapolation of data over a wide energy range has been shown to be extremely successful in reducing the low energy uncertainties. The ¹⁰B(p, α)⁷Be the reaction cross section is characterized by broad resonant and non-resonant components, where the extrapolation requires a careful analysis of the interference patterns in the multi-channel approach [209]. Characteristic low energy pattern and the deviation of accelerator based ⁷Be yield and shot based ⁷Be yield may serve as signature for mapping plasma screening effects! The relatively long half-life of T_1/2 = 53 days makes it an interesting case for diagnostic use by activity measurement. The identification of interference patterns could provide an excellent signature that can be used to extract the plasma conditions during the shot period.

Zoom In Zoom Out Reset image size

Nuclear reactions with low energy light ions (1–50 keV) and atomic targets with bound electrons, show a cross section enhancement from a lowered Coulomb barrier due to screening compared to a bare nucleus target. Enhancements due to electronic screening effects are expected to be present in stellar plasmas as well. The screening effects increase with decreasing reaction (Gamow) energies but are small and a quantification requires a high accuracy (>10%) for the cross section and S-factor measurements.

The DT fusion reaction results in the emission of a neutron and an alpha particle (⁴He), that is, D + T → n + ⁴He, with the release of 17.6 MeV of energy in the form of kinetic energy of the neutron (14.1 MeV) and the alpha particle (3.5 MeV) [220]. The ICF concept as described above (section 1) generates high temperatures and pressures creating a HEDP in which the fusion takes place. The spectrum of neutrons escaping from the ICF capsule is a rich indicator of the conditions and properties of the hot dense plasma and its environment. In contrast, most of the alphas that are created in the ICF implosion do not escape from the tiny hot core, and those that do are quickly stopped in the cooler, less dense fuel plasma immediately adjacent to the core. Nevertheless, these alphas play a crucial role in the working of an ICF capsule. The alpha particles, born with 3.5 MeV of energy, return that energy to the plasma core and the nearby cooler fuel as they slow down, continue to heat the core and fuel, and help create the burn front (and burn wave) that can lead to ignition in future ICF designs (see, e.g., [221]). Despite their importance, there are almost no direct measurements on or diagnostics of these alpha particles in HEDPs. Their travel distance in the ICF capsule is just far too short to effectively detect them by external diagnostics.

A critical, and essentially unmeasured, aspect of alpha particles is their stopping power in a HEDP. Recent theoretical investigations of alpha particles in plasmas have been carried out by a number of investigators [151, 156, 222]. Though all of these models share the same foundational physics to a large extent, the choice of terms included in the models, and approximations to express these terms, lead to differences in calculated alpha stopping power in a HEDP, and, hence, to travel distance, to energy loss, and to distribution of energy loss among electrons and ions. Recently a key step towards ignition in an ICF capsule was taken for the first time [223] in experiments (the 'high foot' campaign) at the NIF where the energy produced by the DT fuel exceeded the energy deposited by the compression process into the fuel and hot spot; the fuel gain was close to 1. A significant part of the improved fuel gain for these shots involved the amount of yield deduced that is attributable to alpha self-heating, that is, the additional yield resulting from the return of 3.5 MeV of alpha kinetic energy to heating the plasma and resulting in more fusion. The ratio of the alpha self-heating yield to the compression yield was almost 1 for one high foot shot. A follow-on phase of the high foot campaign [224] replaced the gold (Au) hohlraum with a DU hohlraum. Significantly, the yield amplification (the ratio of the sum of the compressive yield and the alpha self-heating yield to the compressive yield) was about 2 for several high foot shots with DU hohlraums, compared to 1.7 with the Au hohlraum. The improved overall performance with the DU hohlraums, mostly due to opacity changes, enabled higher implosion velocity, more compression, and rounder implosions [224]. The inferred alpha self-heating and yield amplification, however, have not been causally tied to these or other measurable quantities.

For the alpha particles, the fraction of energy deposited in the hot plasma core, f_α, was calculated to be in the range from just above 60% to 80% for several of the high foot shots [223]. A separate analysis [225] calculated that for a high foot shot f_α was about 56% and the fraction of alpha particles absorbed in the small hot core was 62%. Inferences such as these, the growing importance of alpha self-heating, and the fact that alpha stopping models have not been experimentally tested, suggest that quantifying alpha particle populations and their spectra in HEDPs would be informative.

Radiochemical activation experiments have the potential to provide useful information on alpha particles in HEDPs. For example, a small amount of target material mixed with a portion or all of the DT fuel could be activated via the (α, n) or (α, p) reaction, collected as a gas or solid, and the product's radioactive emissions counted to determine the number of reactions that take place [226]. Another class of experiment that exploits the short range of the alpha particle is a mix experiment where the detector material is added to the capsule shell, either uniformly or as a thin layer on or near the inside surface. Activation of this material could indicate a deep mixing of the capsule material through the bulk of the cold fuel to just outside, or even into, the hot spot [227].

The rest of this section concentrates on the possibility to use the ¹⁰B(α, n)¹³N reaction to assay alpha particles in or near the small hot core of the implosion. The (α, n) reaction with ¹⁰B has suitable nuclear reaction energetics (see table 4), and its cross section increases up to about 25 mb as the alpha energy increases toward 3.5 MeV, enabling the ¹³N product nuclei to be made in adequate numbers. In addition, ¹³N is a gas, and assuming the plasma chemistry produces mainly N₂, can be collected with efficiencies approaching 100% with the RAGS system [6, 228]. The ¹³N half-life is a convenient 10 min. When needed, ¹⁵N could be used as a tracer gas. A drawback of ¹³N is that it decays by positron emission resulting in two 511 keV photons, which is also a typical decay mode for other light gaseous reaction products that might be formed at NIF. Fortunately, the half-lives of these other products are sufficiently different from that of ¹³N that the 511 keV photo-peak decay curve can be successfully de-convoluted to extract the portion due to ¹³N.

The RAGS system was developed to collect, separate, purify, and prepare gas samples for in situ gamma-ray counting or transfer to a gas bottle for remote assay [6, 228]. It has demonstrated the collection of fission noble gases with half-lives ranging from 9 s to 2.8 h, as well as ¹⁹O (27 s), ¹⁸F (110 min), and ²³Ne (34 s) at detector locations next to system getters, traps, and chambers. The ¹⁸F radioactivity is used to characterize the pumping speed for gases from the TC to the RAGS system; the time to pump half of the remaining gas is found to be 180 s. Recently ¹³N, believed to be from (D, n) reactions on the carbon capsule shell, has been collected on the hot getters, and observed with RAGS system detectors [185].

Several radioactive gases that decay via positron emission are made from isotopes that are often part of, or near, the ICF target. Some are made by the (n, 2n) reaction, while others may be produced by deuteron or proton reactions on materials in the capsule. ¹³N radioactivity occupies a 'sweet spot' between those that have half-lives of about 2 min or less, and ¹⁸F which has a half-life of nearly 2 h and thus can be distinguished from most interfering activities. Particularly troublesome are those materials and reactions that produce ¹³N from isotopes other that ¹⁰B, as they are indistinguishable from the desired ¹⁰B (α, n)¹³N reaction product. Plastic (CH) shell capsules are often used for ICF experiments and higher energy deuterons from the plasma or high energy protons generated in the plastic by neutron scattering of hydrogen [185] can produce ¹³N. Thus plastic ablator capsules cannot be used in experiments using boron; beryllium or glass capsules should be suitable. The estimated production of ¹³N from residual ¹⁴N in the TC is low; this is supported by the measurements on shots without CH capsules [185].

Several methods have been devised for loading radiochemical detector material into ICF capsules, and might be used to incorporate boron into capsules using these methods. For example, two layers of the element xenon, each with a different isotope, have been incorporated into capsules for ρR studies [228]. Other materials, such as silicon and germanium, are regularly incorporated into plastic capsules to provide x-ray emitting layers for a variety of x-ray diagnostics. Other ideas, such as aerogel foam layers to hold a material, and direct chemical deposition of materials on the inner surface of the capsule, are being explored. Direct incorporation of gases species (such as di-borane (B₂H₆)) into the DT gas fill may be possible in some circumstances. Effort will need to be invested in any of these capsules to assure that they can be fabricated, and models will have to help assess if the placement and quantity of boron can be characterized at the time of maximum fusion burn.

Finally, execution of the experiment with the RAGS system to utilize the ¹⁰B(α, n)¹³N reaction to diagnose the alpha particles from DT fusion involves a series of straightforward steps: (1) a capsule with ¹⁰B is placed in the TC; (2) fusion produces alpha particles that react with boron to produce ¹³N; (3) ¹⁵N tracer is injected into the TC; (4) the ¹³N is assayed in situ at the appropriate RAGS system component; (5) all the nitrogen is transferred to a removable transfer sample bottle; (6) the ¹⁵N in the sample bottle is assayed by mass spectrometry to get the nitrogen collection fraction; and (7) the ¹⁵N results and the counting results for ¹³N are used to deduce the number ¹⁰B(α, n)¹³N reactions that occurred in the capsule.

Several research and development activities are important to performance of this (α, n) experiment. The successful fabrication of a capsule with boron appropriately incorporated is required. If di-borane is used, safe fabrication methods can be established because of the toxicity of this gas. The degree to which injected ¹⁵N gas is a robust tracer of ¹³N produced in the NIF capsule must be tested. The RAGS system should have modest upgrades to maximize trapping efficiency of ¹³N. A modern cross section estimate for the ¹⁰B (α, n)¹³N reaction is also needed. In addition, a successful experimental campaign will demand a substantial modeling and computational effort to extract physics knowledge and relevance to the NIF program.

The DT fusion alpha particle plays an essential role in the successful ignition of an ICF plasma. Experiments that obtain measured data concerning the DT fusion alphas will open an exciting window into the internal working of an ICF capsule.

Experiments at NIF offer the unique capability to study neutron-induced nuclear cross sections on nuclei in a high-energy density plasma environment [230–232]. In s-process stellar plasma environments, low-lying short-lived excited states can thermally be populated varying their spin (J) and parity. This may significantly alter the neutron absorption rate in branching point nuclei where such rates are comparable to their beta-decay rates, leading to altered population of predicted universal isotopic abundances [233–236]. Furthermore, NEEC [237–240] and nuclear excitation by electronic transition (NEET) [241–243] processes can occur on the highly-excited states produced by neutron absorption reactions before gamma emission, effectively 'hijacking' an (n,γ) reaction midway through completion [244, 245]. There are currently no existing capture rate measurements on plasma-excited nuclei, for any isotope. In addition, there are no measurements of nuclear reactions on non-isomeric excited states, as typical state lifetimes range from nanoseconds to picoseconds.

Interactions with neutrons in the plasma populate excited nuclear states. At NIF the dense environment and short fusion burn produces a high neutron flux, offering the opportunity to study rapid multi-step reactions impossible at any other laboratory. Current non-igniting NIF implosions produce 10¹⁵ neutrons over less than 100 ps in a sphere only 30 μm in radius. This neutron flux of 10²⁹ n cm⁻² s⁻¹ is very high (exceeding that in supernovae) and is expected to be 10³³ n cm⁻² s⁻¹ in igniting capsules. The combination of this immense neutron flux with high electron and photon density offers to study the atomic–nuclear interactions, such as NEEC and NEET, on short-lived, highly-excited states, with lifetimes on order of picoseconds.

An example of how nuclear processes can be affected in a NIF plasma is shown in figure 65, using ¹⁷⁵Lu as an example 'seed' nucleus. A highly-excited nucleus may be created by (n, 2n) reactions from an approximate Gaussian energy distribution of fusion neutrons. The neutrons are centered at 14.03 MeV with a FWHM of about 400 keV, as produced in a typical non-igniting NIF capsule. This populates a range of energy levels in ¹⁷⁴Lu just below the neutron binding energy, shown schematically in figure 65. In a non-plasma room temperature environment (e.g., an external activation foil), the intermediary nucleus de-excites by emitting a series of gamma rays. However, in a high-energy density plasma, electron and photon interactions will both excite and stimulate de-excitation at rates comparable to or greater than the normal decay lifetime. This broadens both the population of spin and energy states of the highly excited nucleus and shifts it upwards in energy due to the high state density gradient. If isomers exist at lower-lying states, decay cascades will populate these states differently due to the altered distribution. Furthermore, if sufficient energy shifting populates states above the neutron separation energy, a neutron may be emitted that was previously energetically forbidden. This effectively transforms the (n, 2n) reaction into an (n, 3n) reaction, producing ¹⁷³Lu.

Zoom In Zoom Out Reset image size

Measuring the effect of the plasma environment on nuclear reactions at NIF may require development and installation of new instrumentation. Long-lived solid products, such as from gold activation, can be collected with solid debris collectors, though increased solid angle capacity is likely needed. Out-of-plasma activation monitor foils are already deployable at several in-chamber locations near the implosion [231]. For the short-lived lutetium example described here, compact in situ gamma-ray detectors are needed, deployed both near solid debris collectors and activation monitor foils shielded from debris. Efforts are currently underway [230] to detect nuclear plasma effects in highly-excited ¹³³Xe gas (produced from ¹³⁴Xe (n, 2n) reactions), which can be collected using the RAGS diagnostic. The control ¹³⁴Xe activation monitor is contained in gas cells [232] located on nearby DIMs.

Two main mechanisms are believed to be responsible for the formation of most of the elements heavier than iron (A = 56); the slow (s-) and rapid (r-) neutron capture processes. A significant amount of effort has been invested in modeling these processes and providing the nuclear physics data (cross sections, masses, β-decay lifetimes etc) needed for their interpretation. Direct measurements of the relevant nuclear physics parameters are difficult-to-impossible to perform since the s- and r-processes involve short-lived radioactive nuclei and take place in HEDP environments with temperatures greater than 20 MK and high particle densities. However, important nuclear physics issues that impact the s- and r-processes can be studied at new upcoming facilities such as NIF, FRIB and FAIR. Experimental data will provide for their incorporation into state-of-the-art astrophysical reaction network models. s-process branch point cross sections: The first of the issues addressed will be the formation of elements in the s-process. The s-process is the best-understood nucleo-synthesis mechanisms, with a weak component (56 < A < 100) acting in the cores of massive stars, and a main component (making A > 100 nuclei) acting in the helium burning shells of 'asymptotic giant branch (AGB)' stars. The s-process has two temperature components (k_BT = 8, 25 keV) and takes place in plasma environments similar. The s-process proceeds slowly via a series of neutron capture reactions between stable isotopes along an isotopic chain until an unstable isotope is formed. If the lifetime of such isotope is long enough (>1 year), a 'branching' can occur where the (n, γ) rate forming the next isotope competes with its β-decay. Table 5 below shows the s-process path in the Tm–Yb region. These 'branch-point nuclei' carry important information on the temperature and neutron environments inside AGB stars. In addition, precise knowledge of the final s-process abundances is essential since they must be subtracted from the measured stellar abundances to determine the contribution by the r- and other processes.

There has been extensive work over the past two decades determining the 'Maxwellian averaged cross sections' for stable nuclei along the s-process pathway, particularly by the group at Karlsruhe [246–248] who have pointed out the need to determine the neutron capture cross sections on branch-point nuclei. However, to date, no capture cross sections have been measured on these nuclei due to the combined effects of the target activity and the low-intensity of neutron beams. A further complication in the extraction of (n, γ) cross sections on s-process branch-point nuclei is the fact that the process occurs in a dense plasma environment believed to induce the population of low-lying excited nuclear states. The latter results in different β-decay lifetime in the stellar environment. Takahashi and Yokoi [249, 250] have compiled tables to reflect these changes in β-decay lifetimes that are commonly used in nucleo-synthesis models. However, there is also significant uncertainty about the effect of populating low-lying excited states on the (n, γ) cross sections. Whereas for the ground states, the lifetimes of such highly-ionized systems can directly be measured in storage rings, the measurements of β-lifetimes of the exited states seems unrealistic at present. The difference in spin between the ground and excited states can be expected to induce sizable changes in low-energy neutron-induced cross sections. These changes in (n, γ) cross sections are generally parameterized in terms of a 'SEF' and are obtained using reaction model calculations that can have significant uncertainties greater than 20%–100%. Therefore, it is of utmost importance to search for experimental capabilities to benchmark the theoretical calculations

The s-process branch points at ¹⁷⁰Tm and ¹⁷¹Tm are examples of a nuclei which could have their (n, γ) cross sections effected by thermal-induced population of low-lying excited states. The ground and first excited states of ¹⁷⁰Tm and ¹⁷¹Tm are separated by 38.71 and 5.03 keV, respectively and have spin differences of 1ħ (see above). These states can be assumed to be strongly populated via nuclear–plasma interactions in the s-process environment with k_BT ≈ 30 keV and densities greater than 100 g cm⁻³.

An ideal approach to improving the modeling of (n, γ) cross sections on s-process branch point nuclei would involve a combination of measurements at accelerators and ICF facilities. Significant progress has been made in the past several years in several accelerator-based techniques to improve the modeling of neutron capture rates using Hauser–Feshbach reaction via the measurements of the nuclear level density (NLD) and radiative strength function (RSF). Larsen and Goriely [251] have shown that even a modest enhancement in the RSF can lead to significant increases in neutron capture rates on neutron-rich Cd and Mo nuclei along the r-process path [252]. These experiments could include particle-gamma coincidence measurements of the NLD and RSF using the Oslo method which have already been used to inform neutron capture rate calculations in neutron-rich ²⁴²Pu [253], and decay-based beta-Oslo decay measurements that have provided similar insight into ⁷⁵Ge(n, γ) [254].

A complementary ICF-based component would involve loading capsules with small quantities (≈10^14–17 atoms) of radioactive target nuclides and collecting the reaction products post-shot to provide energy-integral validation of the capture cross section that include the effects of spin change due to plasma-induced excitations. Experiments to probe these effects are already underway at the NIF where nuclear–plasma interaction induced changes in the population of the ¹³³Xe metastable state. Such experiments are ideally suited to DD and HT-fuel (proton + triton) loaded capsules that produces neutron energies that are more compatible with astrophysical reactions. The approach would involves implanting small quantities of a s-process branch point nucleus together with a neighboring stable isotope where the (n, γ) cross section has been measured in the stellar thermal energy region (k_BT ≈ 30 keV) into the NIF capsule and measuring the production of the radioactive (n, γ) reaction products post-shot. The resulting ratio of capture products allows for the determination of the energy-integrated s-process branch point neutron capture cross section relative to the known stable isotope cross section. State-of-the-art simulations indicate that more than 85% of the neutron captures that occur in the NIF capsule will be due to near-stellar thermal energy neutrons (k_BT ≈ 10 keV). Furthermore, the shot-integrated neutron fluence experienced by a nuclide seeded into an ICF capsule, which is of the order of 10¹⁹ n cm⁻², is comparable to what would be found in the interior of the AGB star (≤10⁷ n s⁻¹ cm⁻²) over the course of more than 10⁴ years. These energy-integrated cross sections include not only neutron capture on the ground state of these nuclei, but also on any low-lying excited states populated via nuclear–plasma interactions.

Table 5 lists the most important s-process branch point nuclei together with the energies of the first excited states. Nuclei likely to have thermal populations of low-lying excited states are indicated in bold face.

The combination of these measurements, e.g., at an accelerator and NIF would then allow a more complete understanding of the s-process and dramatically improve our ability to use the s-process to understand the mechanics and thermodynamics of the interiors of AGB stars.

4.3.4.1. r-process fission recycling

Investigations of nuclear physics processes in a HEDP environment is of utmost importance for our understanding of nucleosynthesis in stars. However, such investigations are highly complicated. It is therefore useful to perform dedicated studies of individual nuclear processes under isolated, e.g., outside of HEDP, conditions. Heavy-ion storage rings offer unparalleled possibilities for a range of such complementary investigations to unravel individual processes that take place in a complex HEDP environment.

4.3.5.1. Storage ring facilities and their properties

Atomic–nuclear interactions in an HEDP environment are predicted [267] to cause excitation of nuclear states with energies comparable to the surrounding plasma. A theoretically predicted NEEC process as depicted in figure 66, has never been observed. Various experiment designs have been suggested of how to verify the process [268]. One scheme is based on bare ²³⁸U ions interacting with an H₂ gas target utilizing the ESR storage ring [269, 270]. This scheme is proposed based on a detailed theoretical evaluation [268, 269] of the ²³⁸U case. In the NEEC process bare ²³⁸U⁹²⁺ ions and free electron collisions result in free electron capture into the L-shell and nuclear excitation of the first excited state at 44.9 keV with subsequent x-ray and γ-ray decay.

Zoom In Zoom Out Reset image size

The theoretical effort has provided the radiative recombination (RR) rates into the L-shell of bare actinide nuclei, ²³⁸U⁹²⁺ and ²³²Th⁹⁰⁺ [268]. In the RR process, the collision of a highly charged ion with a continuum electron can lead to an excited nuclear level via NEEC. In addition, NEEC can occur via an excited intermediate electronic state with the simultaneous excitation of a low-lying nuclear level, followed by fast x-ray emission, NEECX [268].

In this process, the atomic decay times are faster than the nuclear decay (NEECX) and thus the intermediate states determine the decay mode. The atomic decay, from the L-shell to the K-shell, results in an open L-shell and IC blocking; since the K-shell electrons cannot internally convert. The lifetimes for decay of one electron L-shell configurations 2s_1/2, 2p_1/2, 2p_3/2 of U⁹¹⁺ are 5 × 10⁻¹⁵ s, 2 × 10⁻¹⁷ s, 3 × 10⁻¹⁷ s, respectively; the resonance widths are about 30 eV [268]. The lifetime of the first excited nuclear state is determined by its photon decay, which is much slower than IC, and as a result the nuclear lifetime is increased considerably. The nuclear lifetimes are about 0.3 ns (neutral) and 185 ns (bare U⁹²⁺), i.e. more than two orders of magnitude longer if the captured electron (L-shell) decays to the K-shell. The nuclear lifetime variation for the capture into the 2p_1/2 and 2p_3/2 vacancies is due to the difference in gamma decay rates, as determined by the IC coefficients α = 129 and 69, respectively. For the multi-step NEECX process the theoretical results for bare ²³⁸U ions show that the integrated cross section (resonance strength) for NEEC followed by γ decay is 5.4 × 10⁻² b eV (2s_1/2), 1.58 b eV (2p_1/2) and 1.16 b eV (2p_3/2). The continuum electron resonance energies (E_C) are 10.783 keV (2s_1/2), 10.706 keV (2p_1/2) and 15.269 keV (2p_3/2).

The predicted increase in lifetime and resonance strength for NEECX into the L-shell of bare U (and Th) ions indicate that these actinides are promising candidates for the experimental verification of NEEC in the ESR. The long nuclear lifetimes enable a clean separation of the nuclear γ signal from the x-ray background signal due to RR. For the first attempt to verify the NEEC process, the larger resonance strength for ²³⁸U and the previously demonstrated high quality bare ²³⁸U⁹²⁺ ion beams, make ²³⁸U a more favorable candidate than ²³²Th.

The experiment can be performed as follows [269]. The bare U⁹²⁺ ions will be produced using a well-established method that utilizes a sequence of accelerators and strippers following the initial U ion production as a low-charged ion in an ion source. A mono-isotopic cooled fully-ionized ²³⁸U⁹²⁺ ions will be stored in the ESR at an energy of 400 MeV/u. The maximum number of U⁹²⁺ ions is expected to be about 5 × 10⁹. The ions will be stochastically cooled. It is estimated that the ion beam momentum spread can be kept at δp/p = 7.5 × 10⁻⁴ [269]. The electron cooler can be used to reduce the beam momentum spread further.

As a target of electrons, the internal H₂ gas-target will be employed. The gas target will be oriented in a crossed beam arrangement and provide a high electron density that relaxes the requirements for gamma detection efficiency of the gammas released from nuclear decay following the NEEC process. However, this adds an uncertainty regarding CX (nuclear Coulomb excitation) contributions to the nuclear excitation. To minimize the excitation via CX the U ions will be decelerated to the lowest feasible beam energy. To fulfill the resonance condition for NEEC, the U⁹²⁺ ions need to be decelerated to ∼20 MeV/u. This is within the working range of the ESR and has been routinely achieved. The achievable ion beam energy spread is taken into account in the rate estimate. The number of ions after the deceleration can safely be assumed to ∼2 × 10⁷ [269].

The schematic view of the experiment is given in figure 67, which shows the position of the gas target and the photon and particle detectors at the ESR. Detectors at the gas target will measure the REC (radiative electron capture) x-ray emission, which will be used for absolute NEEC yield normalization since the K-shell REC cross section is well known in this regime [271]. The NEECX process occurs simultaneous to the L-shell REC. This process is the non-resonant capture process resulting in the same initial and final atomic states as the NEEC process; however, the REC yield exceeds NEEC by orders of magnitude (the NEEC signal to background is about 10⁻³); discriminating angular dependences are experimentally unfeasible at this point. A good signal to background ratio can be obtained by taking advantage of the different time scales at which NEECX and REC processes occur. In addition, the background x-rays will directly be measured during the experimental campaign. A major advantage for all of the planned measurements is the extremely high detection efficiency for charge-changed ions in the storage ring [272]: all ions, which have captured an electron, continue almost undisturbed in their flight through the ring and will be separated by a dipole magnet from the main beam. A particle detector placed in (or after) the bending magnet will detect them with a near 100% efficiency. Then, coincident counting may be applied to further reduce unwanted background.

Zoom In Zoom Out Reset image size

The gas target will provide a high target density of up to 10¹⁴ H₂ cm⁻². However, it will also include the probability to excite the first low-lying nuclear excited state via CX. The probability for the direct CX to the first excited state is very small as compared to being fed via a cascade of γ-rays from higher-lying states. An estimate of the probability for the indirect CX excitation at 20 MeV/u²³⁸U beam on hydrogen can be done by scaling from a previous experimental work by McGowan [273, 274], which observed thick target CX of ²³⁸U and ²³²Th using an 18 MeV α-particle beam and a Ge(Li) detector.

The work by McGowan and Milner showed that the direct CX excitation rate into the first excited state was approximately 4 excitations/nano-Coulomb as compared to a cascade excitation rate of 1000/nano-Coulomb [273, 274]. If the adiabaticity of the collision scales as e^−2πη with η = α h v ΔE sin (θ/2), then the scaled cross section is less than 0.1 mb for 20 MeV protons incident on a ²³⁸U target. By assuming 5 × 10¹³ target atoms cm⁻² and a repetition rate in the ring of 5 × 10⁵ Hz, this would result in about 0.4 excitations per 2 × 10⁷ particles in the ion train s⁻¹. Since 99.6% of this CX is a result of cascades from high lying states, detection of these high-energy γ-rays can be used as a normalization for CX to the first excited state. These higher-lying states will decay promptly and the corresponding gammas will not contribute to the count rate on the detectors 3.5 m downstream the target. The Compton profile of the H₂ target ions will be Doppler broadened to about 800 eV and needs to be taken into account for computing the resonance width (10–20 eV) and the rate estimates accordingly. Due to the wide energy spread of the observed photons a stringent detector energy resolution is not required.

The rate of NEEC events for the gas target can be estimated from R_NEEC = S N_e N₁ f $[1/{({{{\rm{\Gamma }}}^{2}}_{{\rm{Compton}}}+{{{\rm{\Gamma }}}^{2}}_{{\rm{Beam}}}+{{{\rm{\Gamma }}}^{2}}_{{\rm{Atom}}})}^{0.5}]$ to be ∼10⁵ d⁻¹, where the resonance strength S = σ/Γ_Atom is a reduced cross section per unit energy, N_e (5 × 10¹³ cm⁻²) is the electron areal density of the target, N_i (2 × 10⁷) is the number of ions per bunch, f (6 × 10⁵ s⁻¹) is the frequency of the bunch, and the term in parentheses accounts for nuclear resonance energy mismatch due to the Compton profile, Γ_Comp (800 eV), energy spread of the ²³⁸U beam, Γ_beam (0.33 eV), and the atomic resonance width of the captured electronic state Γ_{Atm_width} (16.5 eV). Thus, a signal from gamma decay after NEEC is expected to be stronger than background due to gammas emitted after CX, ∼3.5 × 10⁴ events d⁻¹.

The number of detected gammas incident on a 5 cm diameter Ge detector located in a detector pocket ∼3.5 m downstream from the gas target and ∼1.5 cm from the beam is given by N_γ,Ge = R_NEEC *Ω_Ge and is about 100 γ d⁻¹, where Ω_Ge is the solid angle of the detector including the detector geometry relative to the beam and the number of decays within the detector viewing window assuming a lifetime of 100 ns.

The NEEC process [268, 275] is an atomic collision process where an initially free-electron is captured resonantly into a bound atomic orbital accompanied by simultaneous excitation of the nucleus, see figure 66. This is the exact inverse of the decay of nuclear states by IC. In addition to the straightforward motivation to study this fundamental coupling between nucleus and electron shells, this process may play an essential role in population of exited nuclear levels in HEDP. The latter is essential for reliable calculation of nucleosynthesis in stars.

One of the main background sources in NEEC searches are events due to Coulomb excitation, which probability depends strongly on the particle energy. Therefore, it is proposed to use the unique feature of CRYRING to slow the ions to lowest energies well below 1 MeV/u [263]. Performing experiments with stored beams and internal targets at such low energies is not easy. Therefore, the present suggestion is based on a possibility to extract low energy beams from CRYRING. Since the beam does not need to be stored in the CRYRING for a long time, it is possible that energies in the order of a few 10 keV/u can be approached.

As a test candidate to search for NEEC, a long-lived isomeric state can be used. For instance, ¹²⁹Sb has a 17.7 min (19/2⁻) isomeric state at 1851 keV. It is noted, that this isomer might happen to be not the best choice, but the experimental procedure will change very little if a different isomeric state is considered in a real experiment. In the chosen example, the next excited state is merely 10 keV above the isomeric state. If one succeeds to induce a NEEC transition from the isomer to this state, it can be identified by an intense (∼100%) 732 keV γ-transition, see figure 68.

Zoom In Zoom Out Reset image size

A unique feature of the existing GSI accelerator family is that a beam of ^129g,mSb can be produced in the FRS, purified from all contaminations and then be injected into the ESR as a pure mono-isotopic beam. The high kinetic energies allow for producing ^129g,mSb as fully-ionized or few-electron ions. In addition, the ultra-high vacuum conditions allow for preserving this charge state for a long time. The ESR is a high-resolution mass spectrometer, which can be used for further purification of an isomeric beam of ^129mSb.

The pure beam of ^129mSb will be slowed down in the ESR to about 4 MeV/u and transported to the CRYRING, where it can further be slowed down to several tens of keV/u. At these energies, the beam of fully-ionized or few-electron ^129mSb ions will be extracted and implanted onto a movable tape. The low kinetic energy has to be chosen to exclude Coulomb excitation of the isomer in the implantation material. The isomers will neutralize by capturing about 50 electrons. If the NEEC process occurs and populates the 1861 keV state, a 732 keV γ-line will be emitted, which can easily be measured by a conventional γ-detector. The neutralization will be accompanied by a huge background of <50 keV x-rays. A suitable absorber can be applied to prevent this background radiation from swamping the detector. The background from spontaneous decay of the isomer can be reduced by moving the tape after each implantation.

The number of possible candidates for the demonstration of NEEC is huge, and since all elements can be studied at GSI the best isomeric state will be selected for the pioneering experiment. Furthermore, if the pioneering experiments are successful, a systematic investigation of the NEEC as well as nuclear excitation in electron transition (NEET) [276, 277] processes and their dependence on the excitation-energy, transition multi-polarities, atomic charge state, atomic number etc can be envisaged.

The resonance properties of the electron shells can provide conditions for the reverse process of nuclear excitation by external electromagnetic fields in HEDPs. Such processes include the time-reversed discrete conversion, also known as NEET and the time-reversed continuum conversion, also known as NEEC. Experimental verification of the NEET and NEEC processes is scarce or not existing (NEEC).

Furthermore, even more exotic decay modes can be searched for, like, for instance Pauli forbidden bound internal conversion (PFBIC), which is proposed by Karpeshin et al [278]. The PFBIC, see figure 69, may take place in ¹⁹⁷Au at ionizations with q > 69+. Li-, He-, H-like ions and bare gold nuclei have 3s electron vacancies and the BIC process turns into Pauli-allowed and all cascade K- and L- x-rays can experimentally be observed. Observation of the PFBIC process would serve for testing fundamental questions of the QED and Fermi statistics in physics [279, 280]. Experimentally one can address this process by utilizing heavy-ion accelerators. For instance, by using a high-Z target foil (e.g. ²³⁸U) where the excited states in ¹⁹⁷Au that decay into its first excited state are produced via Coulomb excitation. The subsequent nuclear γ-decay would be observable down-stream together (if exists) with the two-photons from the PFBIC atomic transitions.

Zoom In Zoom Out Reset image size

If the PFBIC process exists, it may shorten lifetimes of nuclear states by orders of magnitude (20 orders in the case of 76 eV isomeric state in ²³⁸U), or may act to this effect as a natural resonator. This refers in particular to the discrete, or bound, or resonance conversion, also known as BIC (or TEEN (inverse to NEET)), when the transition strength is within a close vicinity of the atomic line. As a result, manipulating with a moderate-power laser, well tuned at the atomic resonance, or in HEDP environments one can attain the prompt decay of isomeric nuclear states, accompanied with the release of energy stored in them.

Nuclear excitations in HEDP are in general of great interest as fundamental physical processes. They affect the formation of heavy elements in stars, and their understanding is thus important for the development of models of galactic chemical evolution and astrophysical processes where they may take place, including the rapid (r-) and slow (s-) neutron capture processes. Moreover, nuclear excitation can be used for nuclear diagnostics of HEDP at facilities such as NIF, ELI, HIPER etc, by placing limits on time-integral temperature and pressure conditions in HEPs.

First results from NIF include the observation of both high-spin isomeric states in ¹⁹⁶Au produced following 14 MeV neutron-induced reactions, and keV neutron energy capture products. The reactions are induced by neutrons from the highly compressed and heated DT fusion plasmas reacting with the Au in the NIF hohlraum indirect drive design. In this case, the Au hohlraum plasma is exposed to the neutrons and a comparison with future experiments, with Au loaded into the cryogenic DT target capsule, will allow to evaluate plasma effects as described above. In this kind of experiments the isomeric excitation is caused by neutrons which serve as a diagnostic for NPI occurring on the highly-excited compound nuclear states populated immediately following emission of evaporation neutrons. The effect of the plasma-induced NEEC/NEET reactions on Au nuclei in the capsule/hohlraum plasma would be to decrease the population of isomeric states produced by 14 MeV neutrons by shifting population away from the high-spin states that would feed the J = 12 isomer to the more abundant low-spin states that are more likely to feed the J = 2 ground state. However, these experiments suffer from limited solid debris collection efficiency would dramatically benefit from the improvements in this area.

At FAIR, the kinetic energies of stored ions in all considered heavy-ion storage rings are much higher than the thermal energies attained at the NIF. However, in NIF experiments, many complex plasma processes occur. By contrast, storage rings offer clean experimental conditions, which might isolate individual nuclear processes. One example is the NEEC process, which remains unobserved despite much research over several decades employing various methods. It should be emphasized though, that although the particles can be highly relativistic in the laboratory frame (E ∼ 400 A MeV in the ESR or E ∼ 5 A GeV in the HESR), the collisions with electrons in the center of mass system are tuned on meV levels. Thus, there are distinct advantages connected to the capability to precisely control the beam velocity, to the use of ultra-thin windowless targets and to having high resolving power. The high energy at HESR and the associated large Doppler shifts will enable laser spectroscopy on atomic transitions that are otherwise inaccessible. This capability suggests many exciting extensions, which will be developed in the future at the FAIR facilities. How it will be possible to employ the HESR, for example, for experiments directly relevant to the NIF will be a task for future research proposals.