Neutron gas scintillation imager with glass capillary plate
Abstract
1. Introduction
2. Neutron gas scintillation with capillary plate (CP)
2.1. Principles of a CPGD for neutron imaging
Fig. 1 shows the operating principle of a CPGD used for neutron imaging. The detector consists of a conversion layer containing 10B and a CP placed in a vessel. The vessel is filled with neon or argon gas mixtures. Charged particles (𝛼-rays and 7Li nuclei) are generated by a nuclear reaction between incident neutrons and the 10B. The charged particles ionize the gas molecules and then generate electrons in the gas. The electrons drift to the surface of the CP along the lines of the electric field and are divided into a number of capillaries.

Fig. 1 Principle of the neutron gas scintillator consisting of a converter and a capillary plate. (a) Proximity-type configuration and (b) direct-mount configuration
In each capillary, the divided electrons multiply and scintillation light is simultaneously emitted by de-excitation of the gas molecules. This scintillation light passes through a glass window and can be observed using a CMOS camera. Fig. 1a shows the proximity configuration used in our previous study. The distance from the converter to the surface of the CP is about 300 μm. In this case, the scintillation light is emitted from many capillaries for a neutron event. Since the image of a charged particle is recorded as an image of its projected track on the surface of the CP, the spatial resolution for a neutron is limited by the track length of the charged particle in this detection scheme.
The configuration that we have proposed to improve the spatial resolution is shown in Fig. 1b. In the configuration, the 10B converter is directly mounted on the inlet surface of the CP. Since the track length of the charged particles is restricted to within the capillary, the generated electrons and scintillation light are produced only within the CP for a neutron incident event. Using this configuration, the spatial resolution of incident neutrons will be improved to values close to the capillary diameter.
2.2. Simulation study
The performance of the detector with the direct-mount configuration as investigated using the simulation code of the particle and heavy ion transport code system (PHITS). The parameters used in the simulations are as follows. The thickness of 3 μm for the 10B layer was chosen to achieve the maximum detection efficiency for a neutron with an energy of 25 meV. The 10B layer was directly mounted on the CP. The diameter and hole pitch of the CP were 50 and 64 μm, respectively, and the CP thickness was 300 μm. The area of the inner hole of the CP was filled with a gas mixture of Ne (90%) + CF4 (10%) at 1 atm. The incident energy of neutron was 25 meV and the distribution of the neutron beam was a pencil source. The total number of neutrons was 107.
Fig. 2 shows the simulation results of the tracks of 𝛼-rays. It is clearly seen that the 𝛼-rays emitted from the 10B layer are restricted to within a capillary.

Fig. 2 Tracks of 𝛼-rays in the direct-mount configuration obtained by PHITS simulation. (a) View of geometry in Fig. 1b and (b) view from top of geometry in Fig. 1b
3. Performance evaluation
3.1. Neutron gas scintillation imager (n-GSI)
A schematic view of the n-GSI with a CP is shown in Fig. 3a. It consists of the CPGD, a mirror, lens optics, an image intensifier unit (Hamamatsu C9016-02), and a CMOS camera (Hamamatsu C13440-
20CU). The CPGD consists of a converter layer and a CP in a sealed-type vessel. The converter consists of a 2-μm-thickness 10B4C layer
sputtered on a silicon substrate. The converter is directly mounted on the CP. The distance between the converter and the surface of the CP was set to close to 0 μm. The effective diameter and the thickness of the CP are 27 mm and 300 μm, respectively.
The diameter and pitch of the capillaries are 50 and 64 μm, respectively. Conductive layers are fabricated on both surfaces of the CP to act as electrodes that generate electric fields for the electron multiplication region. The CP detector is filled with a gas mixture of Ne (90%) + CF4 (10%) at 1 atm. The applied voltage between the two electrodes of the CP is 580 V for the gas mixture. The lens optics are used to focus the output image of the CPGD onto the image intensifier unit, which intensifies the image sufficiently for the CMOS to record an image.
A performance test of neutron imaging was carried out at beamline CN-3 of the Kyoto University Reactor (KUR). This beamline was equipped supermirror guide tube. The neutron energy at maximum neutron intensity is about 20 meV and the effective energy range is from 5 meV to 36 meV. The beam size was 2 cm × 9 cm. The neutron fluxes at the exit of the beamline were 7.6 × 105 and 3.8 × 106 n/cm2/s for 1 and 5 MW KUR operation, respectively. The n-GSI was placed 30 cm downstream from the exit of the beamline. The n-GSI with the proximity configuration was also tested for comparison. The distance between the converter and the surface of the CP was set to 300 μm.
3.2. Signal of single event
Fig. 4a and b show the neutron transmission images obtained with n-GSIs with the proximity and direct-mount configurations for the sample of a Japanese character made of Gd shown in Fig. 4. These images were obtained with a 5 ms exposure time of the CMOS camera. The pixel size of the CMOS image is 2048 × 2048 and the area of each pixel is 4.7 μm × 4.7 μm. The entire area of the images corresponds to 9.7 mm × 9.7 mm.

Fig. 4 Neutron transmission images of a Gd sample in the shape of a Japanese character obtained by n-GSIs with the proximity (a) and direct-mount (b) configurations. (c) and (d) Enlarged images of the elliptical and circled areas marked in (a) and (b), respectively
The tracks of charged particles with various lengths are clearly seen in the image obtained with the proximity configuration. On the other hand, there are several bright spots and no tracks of charged particles in the image obtained with the direct-mount n-GSI. Fig. 4c and d show enlarged images of the marked areas in Fig. 4a and b, respectively.
Fig. 5 shows the line profile of one of the bright spots obtained with the direct-mount configuration. The solid line shows the result of fitting with a Gaussian function. The full width at half maximum (FWHM) of the peak at 39 μm was obtained for the line profile of the single spot and is close to the diameter of the capillary of 50 μm. This result indicates that the track length of the charged particle due to a single neutron event is restricted to within a capillary and that scintillation light is emitted from a single capillary for the direct-mount n-GSI.

Fig. 5 Line profile of one of the single event images recorded by n-GSI with the direct-mount configuration shown in Fig. 4b
Considering the number of bright spots (10) in Fig. 4b, the open area of the Gd sample (8.7 mm2), the exposure time (5 ms), and the neutron flux (7.6 × 105), the estimated detection efficiency is approximately 2%. This value is in good agreement with that of 3% obtained by PHITS simulation reported in Ref.
3.3. Spatial resolution
An image of a test chart was taken with the direct-mount n-GSI to
evaluate the spatial resolution. The test chart of the line pair is shown in Fig. 6a. It was fabricated using a 3-μm-thick Gd metal layer on a fused glass sheet of 90 mm × 90 mm size and 1 mm thickness.
The Gd test chart was placed 8.6 mm away from the converter layer of the n-GSI. The neutron flux was 3.8 × 106 n/cm2/s in 5 MW KUR operation. Fig. 6b shows the neutron transmission image of the square region in the Gd test chart obtained with an exposure time of 600 s.
The spatial resolution is represented as the minimum line width capable of discriminating line pairs. Projections of the line profiles within the solid and dashed boxes in Fig. 6b are shown as the solid and dashed lines in Fig. 6c, respectively. From Fig. 6c, the minimum line width at which peaks and valleys of the line pair can be clearly distinguished was 80 μm. The effective spatial resolution of the n-GSI in this setting was evaluated to be approximately 80 μm from these results.

Fig. 6 (a) Photograph of the Gd test chart developed by Segawa et al. (b) Neutron transmission image of the Gd test chart obtained with the direct-mount n-GSI. (c) Solid and dashed lines are projections of line profiles for the solid and dashed boxes in Fig. 6b
4. Conclusion
We have developed a neutron gas scintillation imager (n-GSI) with a capillary plate (CP) that offers a high spatial resolution. It consists of a converter layer of 10B4C, a capillary plate gas detector (CPGD), a mirror, lens optics, an imaging intensifier unit, and a CMOS camera.
To improve the spatial resolution, we investigated a new CPGD that the distance between the converter and the surface of the CP was set to close to 0 μm. Its performance was investigated using a thermal neutron beam at the Kyoto University Reactor. A spatial resolution of 80 μm was obtained in the performance test using a Gd test chart. We are currently developing a n-GSI with fine-pitch and small-hole diameter CP to acquire a fine neutron image.