Beamed neutron emission driven by laser accelerated light ions

We report on the experimental observation of beam-like neutron emission with peak flux of the order of 10^9 n/sr, from light nuclei reactions in a pitcher-catcher scenario, by employing MeV ions driven by high power laser. The spatial profile of the neutron beam, fully captured for the first time by employing a CR39 nuclear track detector, shows a FWHM divergence angle of 70 degrees, with a peak flux nearly an order of magnitude higher than the isotropic component elsewhere. The observed beamed flux of neutrons is highly favourable for a wide range of applications, and indeed for further transport and moderation to thermal energies. A systematic study employing various combinations of pitcher-catcher materials indicates the dominant reactions being d(p, n+p)^1H and d(d,n)^3He. Albeit insufficient cross-section data are available for modelling, the observed anisotropy in the neutrons' spatial and spectral profiles are most likely related to the directionality and high energy of the projectile ions.


Introduction
Neutrons provide many opportunities for the probing of materials in ways which charged particles and ionising radiation cannot. In this context an ultrashort, directional burst of neutrons with high flux in the MeV range would have wide ranging applications. Exciting opportunities for ultrafast studies lie in the area of materials for fusion energy research due to the growing interest in understanding neutron-induced damage at the atomic scale [1]. Furthermore, a compact source of pulsed, MeV neutrons would provide novel capabilities for interrogation of large cargo containers by fast neutron radiography techniques [2,3], where the nature and location of the threat can be identified by simultaneously measuring scattered neutrons and time of flight of the induced gamma radiation.
Significant attention has been paid recently to laser driven sources capable of producing short neutron bursts, and having potential advantages in terms of cost reduction and compactness, reduction of radioactive pollution and ability of radiation confinement by close-coupled experiments. Although a different approach (high energy deuteron-breakup) has recently been reported [4], the most established route to create a laser based neutron source is by employing laser accelerated ions in either fusion or spallation reactions. Since spallation of heavy atoms requires high energy projectile ions, reactions based on low atomic mass nuclei, such as protons, deuterons, lithium ions, are particularly relevant. The neutron yield from nuclear reactions scales with the product of the densities of the interacting species and the cross-section σ, which for most common reactions reaches high values at ∼MeV centre-of-mass (c-o-m) energy. Producing high fluxes of MeV ions using intense lasers is currently a very active area of research. Where a number of emerging ion acceleration mechanisms, such as radiation pressure acceleration [5] and breakout afterburner [6], hold the promise for producing higher energy ions with higher efficiency, target normal sheath acceleration (TNSA) [7] is a well-established and robust mechanism which produces MeV ions with high flux and narrow divergence. Such beams can be readily deployed in a pitcher-catcher setting for neutron generation via nuclear reactions.
In addition to the benefit of high reaction cross-sections, anisotropy in neutron emission is another facet of beam-nuclear reactions. Simulations [8] show that using several MeV ions in the above mentioned reaction involving low Z materials yields a neutron flux strongly peaked along the ion beam forward direction, and that the anisotropy grows further with increase in ion beam energy. A beamed neutron source is highly favourable not only for the aforementioned range of applications, but also for its further transport and efficient moderation to thermal energies, as required for another range of applications [9], including for instance, Boron Neutron Capture Therapy [10]. Whilst some degree of neutron beam anisotropy has been recently reported at low neutron fluxes [11,12], here we show, for the first time, direct imaging of a true beam-like neutron emission with peak flux of the order of 10 9 n/sr. In particular, a neutron beam with a FWHM of   70 10 ( ) and peak flux of (5 ± 2)×10 8 n/sr was captured in nuclear track detectors kept in the proximity of the source. The neutron beam was produced by d(p, n+p) 1 H and d(d,n) 3 He reactions driven in a beam-catcher scenario, by employing protons and deuterons from thin deuterated plastic (CD) foils irradiated by sub-petawatt laser pulses.

Experimental setup
The experiment was performed using the petawatt arm of the Vulcan laser at the Central Laser Facility of STFC, UK. The linearly polarised laser pulse of 1053 nm wavelength, delivering ∼200 J on target after reflection from a plasma mirror, was focused down to a focal spot of ∼6 μm FWHM, providing peak intensity on the target of3 10 20 W cm −2 . Various targets were irradiated by the laser in order to generate energetic ions via the TNSA mechanism, namely gold foils (10 mm thick), 98% deuterated polyethylene C D n 2 4 ( ) (henceforth called CD) foils with and without few microns thick Al foil at the rear side. The ion beams were diagnosed in the earlier part of the experiment using a high resolution Thomson Parabola Spectrometer (TPS) [13] with image plate detectors along the target normal direction, as shown schematically in figure 1. Since traces for ions with the same charge to mass ratio overlap in the TPS, we implemented the differential filtering technique described by Alejo et al [14] in order to extract the deuteron spectra from the diagnostic. ∼2 mm thick solid blocks of CD and graphite were placed ∼3 mm behind the pitcher target (henceforth called catcher) in order to generate neutrons from nuclear reactions. The transverse size of the catcher was large enough to capture the entire ion beam. A full suite of neutron diagnostics was deployed in order to diagnose the spatial and spectral profiles of the neutrons generated in different shots. In order to capture the flux profile of the emitted neutrons over a large solid angle, and due to its low detection efficiency, of the order of 10 −4 , the CR39 nuclear track detector was placed in close proximity (7 mm) to the catcher. The CR39 was shielded by 4.5 mm thick lead in order to stop the high energy protons (up to 50 MeV [15]), produced by the pitcher target, from reaching the CR39. The absolute neutron flux was obtained from the CR39 by using the etching method and calibration given by Frenje et al [16]. Absolutely calibrated Bubble Detector Spectrometers (BDSs) [17] and nuclear activation diagnostics [18] were used, behind the CR39 and along the ion beam forward direction, at a distance of 10 cm and 50 cm respectively from the pitcher target. Since the BDS provides absolute neutron flux in six discrete energy intervals within the energy range from 0.01 to 20 MeV, it is possible to ascertain the flux of MeV neutrons generated in the catcher by discounting the large signal produced by the lower energy, scattered neutrons bouncing within the target area and hitting the detector several times over a period of time. Using 1 mm thick pure indium foils and measuring the decay of 115 In and 116 In, the neutron flux was estimated within two energy intervals, namely 0-3 MeV and 0.7-15 MeV. Six neutron time-of-flight (nToF) detectors consisting of EJ410 plastic scintillators, optically coupled to XP3330 photomultiplier tubes, were used to provide spectral information at different emission angles by the time-of-flight method. The detectors were shielded appropriately (typically by 5-10 cm of lead on every side and 5 cm of plastic all around except the front of the detector) in order to suppress saturation of signal from Bremsstrahlung radiation and to reduce the noise in the signal due to scattered low energy neutrons reaching the detector. The angle of observation and distance of each nToF detector is given in figure 1. The nToF detectors were cross-calibrated [19] against the spectra obtained from the BDS, which were absolutely calibrated by the company[17].

Results
As it can be clearly seen in figure 2(a), the CR39 shows a beamed neutron emission from the catcher along the ion beam forward direction. The (0,0) coordinate in this image represents the ion beam axis, which is near the bottom right corner of the CR39. For comparison one can see figure 2(c) showing the CR39 from the reference shot taken without the catcher, while keeping everything else the same. Both shots had identical pitcher targets of 10 μm CD and measured laser energy to within 1% of each other. As expected, the number of neutrons generated in the latter case (due to the interaction of ions, electrons and gamma rays produced from the pitcher target with the surrounding objects, including the lead shielding of the CR39) was not significant and lower than the detection threshold of the CR39. Neutrons are diagnosed in the CR39 due to the latent tracks created by the knock-on protons, which are revealed after etching in an alkaline solution. The etched pits produced in the CR39 can be seen in the zoomed view of the CR39s, shown adjacent to the respective images. Using the calibration for detection efficiency of CR39 given by Frenje et al [16], the pit density across the CR39 was converted into neutron flux, as plotted in figure 2(e). The horizontal and vertical lineouts of neutron flux across the CR39, passing through the coordinate (0,0), show an axisymmetric neutron flux profile with a FWHM divergence of   70 10 ( ) . The peak neutron flux along the ion beam axis was estimated as 5 2 10 8 ( ) n/sr. In order to identify the dominant nuclear reactions producing the beamed flux of forwardly directed neutrons, a systematic study was carried out by varying different pitcher and catcher materials. As can be seen in figure 3, the pitcher-only configuration using 10 μm CD target generates a fairly isotropic neutron emission in 4π with an average flux~5 10 7 n/sr. This corresponds to a total neutron yield in excess of 10 8 neutrons (>10 6 neutrons/Joule), which is in line with the trend of neutron yield against the incident laser intensity reported in literature (see figure 4(a)). In this case the neutrons are generated either by the thermonuclear reactions in the hot dense plasma produced by the laser interaction, or by fusion reactions in the target bulk driven by the ions accelerated at the laser front surface through ponderomotive processes [28][29][30][31].
The isotropic neutron emission from pitcher-only targets, which is commonly observed in experiments as shown in figure 4(b), is in stark contrast to the forwardly directed, beamed neutron flux obtained in the pitchercatcher configuration as shown in figures 3 and 5(a). As can be seen in figure 4(b), our data represents the highest degree of beam anisotropy observed experimentally with peak flux of the order of 10 9 n/sr. The order of magnitude increase in neutron flux along the ion beam axis (which is the same as the incident laser axis) observed in this case originates from several different nuclear reactions inside the catcher. Possible reactions include the interaction of protons, deuterons and carbon ions, produced from the pitcher target by the TNSA mechanism, with the carbon and deuterium atoms in the catcher. We note that, in contrast with previous results Figure 2. The neutron beam can be clearly seen in the lower right of the CR39 slide in (a). A close-up of the pits created on the CR39 can be seen in (b), obtained by using a transmission microscope with 10x zoom, which can be compared with the images in (c) and (d), taken in a shot without a catcher target. The laser axis was 1 cm up from the bottom of the CR39, and 1 cm in from the right hand edge. (e) shows the horizontal and vertical beam profiles from the scanned CR39 slide, along with the neutron flux (between 2.5-10 MeV) obtained from the nToF and bubble detector data from the same shot.
at lower laser intensities [11], we report here a significantly higher flux in the pitcher-target configuration than for pitcher only targets, mainly due to the much more efficient TNSA acceleration with high energy, PW pulses.
Different pitcher-catcher combinations were used in order to unfold the contributions from different potential reactions towards the observed beamed flux of neutrons. Firstly, employing a graphite catcher with a CD pitcher target (CD  C) did not produce (in spite of shot-to-shot fluctuations) a significant increase in the neutron flux or beam anisotropy compared to a pitcher-only target. This suggests an insignificant contribution being made by the p+C, d+C and C+C reactions in the context of the observed neutron flux in the CD  CD case. Secondly, in order to discriminate between proton and deuteron induced reactions in the CD targets, several shots were taken by using a CD catcher in front of a deuterium free ion source (No _ d  CD) by using either 10 μm gold or CH foils as pitcher targets (Au/CH  CD), or 10 μm CD foil backed by few microns thick aluminium foil as pitcher targets (CD+Al  CD). Similar neutron fluxes as in the CD  CD case were observed in this case (No _ d  CD). Since the deuteron and carbon ion spectra produced from the pitcher targets are similar in terms of number of particles and beam temperature, as shown in figure 6(a), d+C reactions in the case of No _ d  CD can also be assumed to be insignificant. Therefore, the most promising reaction in this case would be the proton driven neutron generation via breakup of deuterons (d(p, n+p) 1 H) in the catcher. This is expected due to (1) the large number of high energy protons (∼20 MeV) being produced from the pitcher target by the TNSA mechanism, as shown in figure 6(a), and (2) the high cross-section of the d(p, n+p) 1 H reaction (although there is a limited amount of data available [32]) as shown in figure 6(b). Since cleaning techniques for removal of target contaminants (such as laser ablation, resistive heating etc) could not be implemented in the experiment due to setup constraints, it was also not possible to study the interaction of a proton-free ion beam with CD catchers. However, by comparing the data obtained from No _ d  CD and CD  CD cases in figure 3, one can reasonably assume similar contributions from d(d,n) 3 He and d(p, n+p) 1 H reactions in the CD  CD case.   [12,[20][21][22][23]), as cited beside the data points. The red star represents the average neutron yield from 10 μm CD target obtained in our experiment, as shown in figure 3. (b) Graph showing the ratio between on and off axis neutron fluxes reported in literature (see [11,12,21,[24][25][26][27]) , as cited beside the data points, with respect to on axis flux obtained in experiments. The data point obtained from our experiment is labelled as [i], [ii] and [iii] which corresponds to different pitcher and catcher combinations, such as CD  CD, Au/ CH  CD and CD+Al  CD respectively. The Au/CH  CD case produced the highest degree of beam anisotropy, which is expected due to the low level of isotropic neutron flux produced from the non-deuterated pitcher targets, as can be seen from figure 3.
Angularly resolved neutron spectra obtained by the six nToF detectors, as shown in figure 5, show the difference in neutron generation between the cases. While the CD  C case produced low energy neutrons isotropically, similar to that obtained for a pitcher only target shown in figure 5(a), No _ d  CD and CD  CD produced a very significant anisotropy in both neutron flux and maximum neutron energy, as shown in figures 5(b) and (c) respectively. Figure 5(c) shows the neutron spectra obtained along different angles for the case shown in figure 2(a), where the spectrally integrated neutron flux over the neutron energy in the range 2.5-10 MeV obtained from the nToF detectors agrees well within the experimental errors with the off axis BDS and CR39 measurements shown in figure 2(e). The highest neutron flux obtained in the experiment was for the CD  CD case, where, as shown in the figure 5(d), the nToF spectra agrees with that obtained from the BDS and activation diagnostics. The peak neutron flux along the beam axis, for neutron energy between 2.5-10 MeV, was close to 10 9 n/sr. Indeed, the neutron flux can be significantly enhanced by optimising the neutron generation with high yield catcher targets, such as lithium or beryllium, for which the reaction cross-section is an order of magnitude higher than for d(d,n) 3 He reaction and stays very high for ions with tens of MeV energy.
The anisotropy ratios (which we define as the ratio of on-axis flux to off-axis flux) that have been observed in these experiments can be estimated for a beam-fusion scenario in terms of the projectile energy and the differential cross-section of the nuclear reaction. For instance, in the case of the d(d,n) 3 He reaction, using the ion  , where E d is the incident deuteron energy in the laboratory (lab) frame, m p is the proton mass and Q is the Q-value of the reaction. Therefore, the neutron energy in the lab frame along a given neutron emission angle (θ, with respect to the incident ion beam axis) can be written as Where, a = v v d,cm n,cm . An anisotropy ratio can thus be obtained at each deuteron energy by taking the ratio of the differential cross-section values that are calculated in this way at 0°and 90°. This requires one to use the tabulated data for the differential cross-section in the c-o-m frame available in the experimental nuclear reaction database (EXFOR) [32].
The results of this calculation are shown in figure 7(a), which indicates that the d(d,n) 3 He reactions could easily produce the levels of anistropy shown in figure 4 given that the deuteron energy spectrum shown in figure 6 extends above 5 MeV. The anisotropy of emission from this calculation assumes that the angular distribution of the incident ion beam is sufficiently narrower than that of the neutrons. Without the differential cross-sections for d(p, n+p) 1 H reactions in the c-o-m frame, it is hard to produce a similar graph for the d(p, n +p) 1 H reaction. However, even if one assumes an isotropic emission of neutrons for the d(p, n+p) 1 H reaction in the c-o-m frame, the kinematic effects still lead to an anisotropy ratio of about 10 over a wide range of energies.
In addition to the observed anisotropy in neutron flux, the energy distribution arising from the reactions, as can be seen from figure 5, is highly anisotropic due to nuclear reaction kinematics. According to equation (1), one can approximate the energy of emitted neutrons along the beam forward direction (see figure 7

Conclusion
In conclusion, we have demonstrated a strongly beamed (~ 70 FWHM), high flux (of the order of 10 9 n/sr) source of fast neutrons based on beam-nuclear reactions employing high power laser driven protons and deuterons. The neutron flux in the beam, which was amongst the highest reported in the literature, was an order of magnitude higher than that present outside of it, as characterised spatially and spectrally by a suite of neutron diagnostics. Such a directed beam of fast neutrons is highly favourable not only for its direct applications, but also for its transport and moderation. With the possibility of producing higher energy ion beams and higher flux, either by TNSA or via emerging ion acceleration mechanisms, and by employing higher yield neutrons converters (such as 7 Li or 9 Be), this approach can lead to the development of an appealing neutron source for both established and innovative applications. Energy of emitted neutrons in a beam-fusion scenario along 0°and 90°to the incident ion beam axis as a function of the incident deuteron energy, calculated using equation (1).