The Bimodal Neutron And Photon Imaging Driven By A Single Electron Linear Accelerator

Both X-ray imaging and neutron imaging are essential methods in non-destructive testing. In this work, a bimodal imaging method combining neutron and X-ray imaging is introduced. The experiment is based on a compact electron accelerator that can simultaneously generate two kinds of radiation: X-ray and neutron. This identication method utilizes the attenuation difference of the two rays’ incidence on the same material to determine the material’s properties based on dual-imaging fusion. It can enhance the identication of the materials from single ray imaging and has the potential for widespread use in on-site, non-destructive testing where metallic materials and non-metallic materials are mixed.

(electron linear accelerator) driven system. The experimental results demonstrate that both the photon image and the neutron image can be acquired simultaneously within the same e-LINAC operation, and be free from the different imaging beam geometries. The pixel-wise matching of the two images can be e ciently conducted to form the inspected object's bivariate histogram to identify different materials of various mass thicknesses. In the case that the inspected sample is evolving with time or stochastic processes, the traditional two-source-detector-system may inevitably introduce error in fusing the two images, because each imaging mode may take several hours or even longer and the inspected sample is perhaps not identical for the two imaging modes. In this study, the time delay between the two imaging modes can be as small as 10 ms (when the e-LINAC works at the repetition rate of 100 Hz), which intrinsically ensure that the inspected sample is identical for the two imaging modes. Consequently, being suitable to inspect samples evolving with time or stochastic processes would be an intrinsic advantage with the one-source-detector-system presented in this study.

Results
System overview In Fig. 1, we illustrate the principle to produce neutrons and photons used for the bimodal imaging driven by a single e-LINAC system. The 9 MeV electrons are very energetic, and hence the bremsstrahlung photons generated on the tungsten target are forward emitted. A heavy water convertor is placed ahead of the tungsten target to generate both the imaging neutrons and photons. Heavy water is chosen as the material to convert bremsstrahlung photons to neutrons due to (1) the low (γ,n) threshold of 2 H (E th = 2.223 MeV) and (2) the superb neutron moderation capability of 2 H and 16 O. The neutron moderation capability is critical in this study because the fast photoneutrons produced by the 2 H(γ,n) 1 H reaction, in general, should be decelerated to slow neutrons to improve the imaging sensitivity. The orbital electrons of the 16 O atoms and those of the 2 H atoms can induce the scattering of bremsstrahlung photons. When the emitting angle of scattered photons is chosen as 90° with respect to the direction of bombarding electrons, the energy of incident photons interrogating the inspected object will typically be less than 511 keV (due to the Compton scattering), as shown in Fig. 2 (a). Photons with such an energy spectrum mainly interact with atoms via Compton scattering, which shows an almost constant mass attenuation coe cient for different elements 13 , and hence are helpful to analyze the mass thickness of the inspected object compared with x-ray tube measurements, as shown in Fig. 2 (b).
The neutron emission direction should be the same as the photons to conform the photon imaging geometry. In fact, the energy spectrum of emitted neutrons is not sensitive to the emission direction due to the almost isotropic moderation process of neutrons within the heavy water converter. In Fig. 3, we present the neutron energy spectrum measured with the time of ight (TOF) method by a 3 He counter placed 10 meters away from the heavy water converter at the angle of 90°. The simulated neutron energy spectrum is also shown, and the two spectra match fairly well. The results show that when the 9 MeV e-LINAC works at 100 µA current, a 2500 neutron/cm 2 /s thermal neutron ux at 10 meters away can be anticipated for the neutron imaging. Its counterpart for photons is 10 8 photon/cm 2 /s.

Imaging sequence
Although imaging photons and neutrons produced by the bremsstrahlung photons share the same imaging geometry, their imaging processes' interference should be considered. As e-LINAC works at a pulse mode of 5 µs duration and 100 Hz repetition rate, the photon ight time from the heavy water converter to the detector is merely 5.033 µs, in which the 5 µs is the pulse width of photons and the 0.033 µs is the photon's ight time across 10 meters. Considering the decay time for the light emitted by the scintillation screen is 0.2 µs, in order to let the photons' in uence on the detection system die away, an additional time delay of 2 µs after the last photon bombarding the n MCP detector should be set for the photon imaging and before triggering the acquisition of neutron imaging. Therefore, in principle, the duration of [7.033 µs, 10 ms] after each electron pulse can be assigned to neutrons for neutron imaging.
In the experiments, we chose [50 µs, 9.95 ms] as the duration for neutron imaging to avoid the mutual interference between the two imaging processes. Thermal neutrons used for neutron imaging have a characteristic speed of 2200 m/s and require 4.5 ms for the 10-meter ight. Thus, both the photon imaging and neutron imaging can be perfectly accommodated by the [33ns, 7.033 µs] and [50 µs, 9.95 ms] durations, respectively, as shown in Fig. 4.
The spatial distributions of imaging neutrons and imaging photons Data acquisition of the last collision positions of neutrons inside the heavy water converter indicates the heavy water converter acts as a volume neutron source. As the detector system is typically placed 10 meters from the heavy water converter, this volume neutron source will be reduced to a surface source with a disk shape (the diameter is 10 cm, determined by the ight tube in Fig. 1(b)), as shown in Fig. 5 (a), with its counterpart for photons shown in Fig. 5  Bene tting from the drastic difference between the attenuation coe cients for neutrons and photons, this technology can help nd the residual core material in cast turbine blade 14 . Fig. 8 (b) and (d) are the photon image and neutron image for a blade shown in Fig. 8 (a) with residual gadolinium tracer (gadolinium oxide powder in this study), respectively, while Fig. 8(c) and (e) are that for a blade without residual gadolinium tracer, respectively. There is no signi cant difference that can be noticed between the Fig. 8 (c) and (e), indicating the inability for photons to investigate the residual gadolinium tracer inside the blade. On the contrary, the difference between Fig. 8 (b) and (d) is evident, implying that the blade with residual gadolinium tracer can be effectively discriminated by neutrons. By fusing the images of Fig.  8 (b) to (e), a new image re ecting the position distribution of residual gadolinium tracer inside the blade is formed and shown in Fig. 9(a). To conduct a more quantitative comparison between the blades with or without gadolinium tracer, the distributions of the value, which is the ratio between the mass attenuation coe cient of neutrons and that of photons, of each pixel in the six squares of Fig. 9(a) are calculated and shown in Fig. 9(b)(1)~(6), with their counterparts for blade without gadolinium tracer are also shown for comparison. Because of the existence of gadolinium tracer, the separation between the two curves in Fig.   9(b)(1)(2)(5) are evident. Due to the lack of gadolinium tracer in Fig. 9(b)(3)(4)(6), the two curves in which conform to each other and does not show signi cant difference. Fig. 9(c)(1)(2) show the bivariate histograms of turbine blades without or with gadolinium tracer. The turbine blade with gadolinium tracer differs from that without gadolinium tracer obviously.

Discussion
The industrial applications of neutron imaging have long suffered from the lack of a suitable neutron source that can deliver an intense neutron beam with a long lifespan 1 . Reactor sources, or spallation neutron sources, are ruled out for their high construction and operating costs. Isotopic neutron sources cannot provide the necessary high-yield neutron beam, and some suffer from the short half-lives 15 . Therefore, only the accelerator-driven neutron sources can be considered. Although the proton or deuteron-induced neutron emission reactions, with lithium or beryllium as the neutron production target, show a relatively larger cross section, their real applications are hindered for two reasons. First, the protons or deuterons that bombard the lithium or beryllium target typically have the 3.5 MeV to 13 MeV energies 16-19 , which limit their penetration depth in the target to around 100 µm 20 . This, in turn, limits the neutron yield since the number of involved target nuclei is small. Second, and perhaps more intrinsically, the target's lifetime is severely affected by the deposited hydrogen elements and the heat generated by more than 10 kW 22 of power deposited within the target's 100-µm thin surface layer. We observed the proton beam broke several beryllium targets after just several hours of bombarding. While the e-LINACdriven neutron source might be considered to possess a lower photoneutron cross section, it can provide the same neutron yield as the proton or deuteron accelerator sources because the large mean free path of the MeV photons, which implies more target nuclei are involved in neutron production. Moreover, both electrons and photons will not deposit particles of non-zero-rest-mass inside the tungsten target and the heavy water, respectively; thus, the source's lifetime can be in nite provided the e-LINAC system operates normally.
Besides the relatively low cost and modest footprint 21 , the most attractive property of the e-LINAC-driven system is that it can provide the imaging photon beam and imaging neutron beam simultaneously with a negligible difference between their imaging beam geometries. The successive photon imaging and neutron imaging measurements within one e-LINAC operation facilitate the fusion of the photon image and the neutron image. This unique property makes the e-LINAC driven system be one of the most promising bimodal imaging systems.
The spectra of photons and neutrons would undergo the hardening process when they penetrate inspected objects of various mass thicknesses. Therefore, the ratio between the neutron attenuation and photon attenuation might not be constant even for a particular material. However, a curve in the plane of Fig. 7(b)and Fig. 9(c) can be anticipated to be assigned to a particular material when its mass thickness is varied. As can be seen in Fig. 10, by improving the counting statistics, the spread of the curve for nickel can be reduced, and the ability for separating different materials can be enhanced.
The most straightforward way to improve the counting statistics is to increase the neutron yield of the neutron source. The neutron yield can be further enhanced when the electron energy, or the current of electrons bombarding the tungsten target, can be augmented. For a 10 MeV/ 20 kW 23 or a 50 MeV/ 25 kW e-LINAC system, the neutron yields can be 40 or 100 times higher, respectively, than the 9 MeV/0.9kW of this study, when their optimal neutron converters are used. The available ux at the detector position can thus be signi cantly improved, and the acquisition times for the neutron image and the photon image can be signi cantly shortened to decades of seconds or less. The high thermal power on the tungsten target deposited by the electrons may pose technical problems similar to other accelerator-based neutron sources. The nite element analysis shows that a water-cooling system can effectively remove the heat deposited on the tungsten target to maintain its robust operation 23 .

Electron linear accelerator
The electron linear accelerator used in this study is a 9 MeV e-LINAC manufactured by NUCTECH Co. Ltd., which can deliver 9 MeV electron pulses of 5-µs width. The repetition rates are adjustable, ranging from 20 Hz to 250 Hz, and the corresponding electron gun current is from 8 µA to 100 µA.

Heavy water converter
Heavy water is used as the photon-to-neutron converter. After the 9 MV bremsstrahlung photons are generated from the 9 MeV electrons bombarding the 1.5-mm-thick tungsten target, photoneutrons are produced when an energetic photon breaks the 2H nucleus into proton and neutron. Monte Carlo simulations were carried out to determine the geometry of the heavy water converter to achieve a high neutron ux of a suitable energy spectrum in the detector position. Total weight of 6.5 kg heavy water is contained within an aluminium vessel of Φ16 cm × 28 cm and 3-mm-thickness.

Neutron moderator
The neutron moderator is made up of high-density polyethylene and heavy water. The generated fast neutrons undergo collisions with the light nuclei in moderators such as 1H, 2H, 12C, and 16O. From the collisions, the velocity of the neutrons will gradually decrease to the thermal neutron region. The optimized outer size of the polyethylene element is Φ36 cm × 44 cm, and the mean energy of the emitted neutrons is 44.26 meV.

Shielding and Collimator
A 10-meter-long vacuum tube is designed for neutron transport and collimation. Eleven ring-shaped neutron absorbers made of boron carbide ceramics are placed inside the vacuum tube with an interval of 1 meter between each to collimate the neutrons traveling inside the tube. The tube's outer surface is covered by boron-containing rubber with 60 wt % boron to absorb the neutrons that may escape from inside the tube. The detector is placed in a shielded container, from inside to outside, by 10 cm lead, 0.5 cm boron-containing rubber, and 30 cm boron-containing polyethylene. The S/B (signal neutrons versus background neutrons, where signal neutrons are the neutrons that travel directly from the heavy water converter, and background neutrons stand for those undergo scattering in the circumstances) ratio of this system is larger than 200.

Detector
Both neutrons and photons are measured by the same detector, a neutron-sensitive micro-channel plate (nMCP) produced by Photonis Co. Ltd. Its sensitive area is 95 mm × 95 mm. The neutron or x-ray will be converted to avalanched electrons by the nMCP at rst. After the absorption of each neutron or photon, the avalanched electrons then produce uorescence on the P46 scintillation screen with a decay time of 200 ns. The pictures present in the P46 scintillation screen will be registered by a CMOS camera (Andor iStar series with image intensi er) with the aid of an optical system that provides a 90° re ection and scaling for matching the sensitive area of the CMOS camera. The shutter of the CMOS camera can be set with a variable time delay with respect to the triggering signal that indicates the production of an X-ray pulse to separate the acquisition of the photon image and the neutron image.

Fusion of images
Bene ting from the almost same imaging beam geometries for neutrons and photons, the pixel-wise matching between the neutron image and photon image can be conducted directly, without additional linear translation, rotation, scaling, or skew. For the same pixel of the inspected object's images, the corresponding values of neutron attenuation and photon attenuation can be extracted from the two images compared with the images of "air" of neutrons or photons, respectively. The value for neutron attenuation and that for photon attenuation of the same pixel determine a point in the coordinate system whose x-axis is the photon attenuation, and y-axis is the neutron attenuation. By analyzing all the pixels present in the two images, a bivariate histogram, in which different materials with various mass thicknesses can be identi ed, is formed.
For both neutrons and photons, the attenuation can be calculated as Where Att is the attenuation of neutrons or photons, equals to the product of μ m and t m , which are the mass attenuation coe cients for neutrons or photons and the mass thickness of the inspected objects, respectively; I 0 and I are the measured counts for neutrons or photons penetrating "air" or the inspected object, respectively. The ratio between for neutrons and photons thus determines a slope to identify different materials given by where subscript n and p stand for neutrons and photons, respectively; the t m cancels because it is the same for neutron penetration and photon penetration. Therefore, when the hardening effect is not severe (the energy spectra of neutrons or photons are not signi cantly in uenced when the inspected object's mass thickness is varied), in the bivariate histogram, a speci c material will have a certain slope. In the case that several different materials are successively penetrated by neutrons and photons, the slope might be modi ed as where N is the number of materials. In general, we cannot separate three or more materials with the information provided by the bivariate histogram. However, considering that the pulse working mode of the e-LINAC-driven system enables energy-resolving neutron imaging, a multivariate histogram can be formed for the further identi cation of materials. We have successfully conducted an isotope identi cation experiment, which realizes the energy-resolving neutron measurement, demonstrating the possibility of the multivariate image data acquisition within the framework of the system in this study.

Declarations Acknowledgements (optional)
This work is supported by the National Natural Science Foundation of China under Grant No. 11735008. We also thank LetPub (www.letpub.com) for its linguistic assistance during the preparation of this manuscript.

Competing interests
Submission of a competing interests statement is required for all content of the journal.

Data Availability
For more information on data availability, click here. Certain data types must be deposited in an appropriate public structured data depository (details are available here) and the accession number(s) provided in the manuscript. Full access is required at acceptance. Should full access to data be required for peer review, authors must provide it. Nature Communications encourage provision of other source data in unstructured public depositories such as Dryad or gshare, or as supplementary information. To maximize data reuse, Nature Communications encourage publication of detailed descriptions of datasets in Scienti c Data.

Code Availability
Research papers using custom computer code will also be asked to ll out a code and software submission checklist that will be made available to editors and reviewers during manuscript assessment.
The aim is to make studies that use such code more reliable by ensuring that all relevant documentation is available and by facilitating testing of software by the reviewers. Further detailed guidance and required documentation at submission and acceptance of the manuscript can be found here. Figure 1 (a) The principle to produce imaging neutrons and photons simultaneously. Energetic electrons are delivered by a 9 MeV e-LINAC and bombard the tungsten target to produce bremsstrahlung photons rstly. The bremsstrahlung photons then hit the 2H nuclei to produce the photoneutrons via the 2H(γ,n)1H reaction and the electrons surrounding the 16O nuclei to produce scattered photons, respectively. Both the neutrons and photons will penetrate the inspected object and be collected by a neutron-sensitive microchannel plate (nMCP) detector and converted to electron clouds, leading to a scintillation on the scintillation screen and formation of the photon and neutron images successively. A complementary metal oxide semiconductor (CMOS) camera then registers the two images with the aid of an optic system composed of the mirror and lenses. (b) Photograph of the bimodal imaging system driven by a 9 MeV e-LINAC.  The simulated and experimentally measured spectra of neutrons along the 90° direction for neutron imaging.     (a) Fused image of the turbine blade; (b) the distributions of μm,n/μm,p for turbine blades with or without gadolinium tracer, for six different positions of the turbine blade (selected region is 40×40 pixels, and each pixel is ~120μm×120μm); (c) the bivariate histograms for turbine blades (1) with gadolinium tracer and (2) without gadolinium tracer. The pixel density stands for the number of pixels within a region of 0.0075 μm,ntm×0.0075 μm,ptm.