Conceptual design of the grazing-incidence focusing small-angle neutron scattering (gif-SANS) instrument at CPHS

1. Introduction

Small-angle neutron scattering (SANS) is one of the most powerful techniques for probing mesoscopic structures of bulk substances. It serves a large community of users from various scientific disciplines. Although SANS instruments have been built in most large neutron research facilities, they are heavily oversubscribed and only about one third proposals for instrument beamtime can be approved. In the past two decades, a surge in the number of compact accelerator-driven neutron sources (CANS) both built and being built is witnessed [1,2]. CANS have been playing a significant role in many applications [1] such as device development, interrogation, nuclear data and neutron capture therapy. However, due to the weak neutron intensity, SANS is not commonly regarded as a viable option for CANS and it consequently has not been widely equipped at CANS. Driven by extensive demands from multidisciplinary studies, efforts have been intensively made to pursue competitive SANS measurements at CANS. These efforts can be roughly categorized into two approaches. The first one is focused on building high-brilliant CANS, such as the ongoing SONATE project [3] and HBS project [4], while the second is devoted to improving the instrument performance with novel devices and methods, including grazing-incidence focusing mirrors [5,6], multi-pinhole collimators [7,8], ring-shaped collimators [9] and so on.

OPEN IN VIEWER

Fig. 1.

Schematic view of the nested supermirror assembly.

Among these neutron optics, grazing-incidence focusing mirrors, capable of bringing orders-of-magnitude increase in neutron intensity on sample by utilizing a large collecting area, are expected to bring competitive SANS capabilities for CANS. Hence, grazing-incidence focusing mirrors have been intensively investigated at the compact pulsed hadron source (CPHS) [10] at Tsinghua University, in the cooperation with Tongji University. Considerable efforts have been paid to achieve a large collecting area through the nested conical optics to simultaneously enabling a nesting structure and supermirror coatings. Finally, we fabricated a focusing supermirror assembly with two nested shells and m = 2 supermirror coatings (see Fig. 1). The performance of this device was measured, and a prototype SANS instrument was built to demonstrate its capabilities [11,12].

Based on the preceding work, we decide, in a bold step, to launch the grazing-incidence focusing SANS (gif-SANS) project aimed at a high neutron intensity on sample ( ⩾ 10 5 n/s) and a large range in scattering vector ( Q min ⩽ 10 − 3 Å − 1 ). The performances of such a gif-SANS can be comparable to the instruments at medium reactor sources, which is significant to pave the way to high-performance SANS at CANS. The preliminary design and calculated performances of the gif-SANS will be presented.

OPEN IN VIEWER

Fig. 2.

Schematic layout of a typical gif-SANS.

Figure 2 displays the basic structure of a typical gif-SANS. The mirrors are segments of ellipsoids, and the detector and the source aperture are placed at the two focal planes of the ellipsoids. Geometrically, we can define a magnification M, as the ratio of the size of the focal spot R 2 and the size of source aperture R 1 . M is a function of the mirror position and is independent of the mirror size [13]. The focused neutron intensity on sample, I f , is proportional to the area of the source aperture, the collecting area of the mirror assembly A m and the wavelength reflectivity R Λ . Incorporating the expression of the minimum accessible scattering vector Q min into the expression of I f leads to the following relation: (1) I f ( λ ) ≃ Φ 0 ( λ ) Δ λ · π ( Q min 2 π / λ ) 2 · A m · ( SDD L ) 2 · ( M 1 + M ) 2 · R Λ ( λ ) where λ is neutron wavelength, I f ( λ ) is the neutron intensity for λ ∈ [ λ , λ + Δ λ ) , Δ λ denotes the wavelength interval, Φ 0 ( λ ) represents source brightness, SDD and L denotes the sample-to-detector distance and the detector-to-source distance, respectively. Equation (1) manifests that achieving a large collecting area and a high wavelength reflectivity are crucial for realizing a high intensity and a small Q min for gif-SANS. The supermirror assembly with a nested structure and supermirror coatings, fabricated by the nested conical optics [12], has been demonstrated to meet the above requirements.

OPEN IN VIEWER

Fig. 3.

Preliminary design of the CPHS gif-SANS.

The preliminary design of the CPHS gif-SANS is shown in Fig. 3. With the layout of the CPHS neutron hall taken into consideration, the instrument length from the source aperture to the detector is set to 8.0 m. The mirror assembly is placed at the middle to keep a symmetric geometry ( M = 1 ). The parameters of the mirror are optimized with numerical calculations and McStas simulations [12] and are shown in Table 1. m = 3 supermirror will be coated to collect as many as neutrons. According to the design philosophy described in Ref. [12], the SDD is set to 3.0 m to avoid the unreflected beam hitting samples and to achieve a smallest possible Q min . The primary detector, consisting of 100 position sensitive ³He tube-detectors with a length of 800 mm and a diameter of 8 mm, will be used to extend Q min down to at least 0.007 Å − 1 (the maximum neutron wavelength is about 9 Å ). Further decreasing Q min can be made by varying the size of the source aperture to shrink the focal spot on detector. This differs from the way the conventional pinhole design uses. For many pinhole SANS, the detector can be moved along the beam axis to adjust the Q-range and covering a wide Q-range often requires several measurements. For the gif-SANS design, the detector is fixed at the focal plane of the mirrors and thus the SDD is fixed. To further achieve a Q min less than 10 − 3 Å − 1 , the radius of the focal spot on detector should be less than 4 mm, which requires a high spatial resolution detector to cope with. Therefore, the CPHS gif-SANS will equip another detector, a neutron sensitive microchannel plate (ⁿMCP) detector with a spatial resolution better than 0.2 mm [14], for lower Q min measurements. To allow for the detector at work sitting at the focal plane, both detectors are movable. As shown in Fig. 3, the ⁿMCP detector can be moved perpendicular to the beam axis, while the ³He tube detectors can be moved along the axis.

A large collecting area is a double-edged sword. It can largely enhance the neutron intensity on sample, but it also requires a large sample area which is not accessible in many cases. To make the instrument more versatile, the pinhole collimation will be incorporated in the instrument design with the help of dedicated mechanical and control apparatus. Another merit of retaining the pinhole option is that it does not filter the short-wavelength neutrons which allows for a large Q max available. Consequently, by using two detectors and two neutron optics, the CPHS gif-SANS have three operating configurations as shown in Table 2. The three operating configurations are termed as low-Q mode, mid-Q mode and high-Q mode, respectively.

OPEN IN VIEWER

Fig. 4.

Neutron intensity and Q min as a function of the source aperture radius.

By varying the size of the source aperture, we can adjust the performances of the gif-SANS. Figure 4 shows the neutron intensity and Q min as a function of the radius of the source aperture that are obtained by McStas simulations. In the simulations, the measured CPHS spectrum is used, the reflectivity of supermirror coatings is a Heaviside step function, and the imperfections of mirrors due to fabrication and alignment are ruled out. It is seen that the intensity and Q min increase upon increasing the radius of the source aperture. For the low-Q mode, the intensity is 7 × 10 3 n/s when Q min reaches 6 × 10 − 4 Å − 1 . For the mid-Q mode, the intensity is about 10 6 n / s when Q min reaches 6 × 10 − 3 Å − 1 . The steps of the intensity are due to the discrete resolution of the He-3 tubes. These results show that the SANS instruments at CANS could be highly competitive.

Yet, there are some practical issues constituting the impediments to the above performances: the use of masks in front of the mirrors reduces the collecting area [12]; the reflectivity of supermirror coatings is less than unity; diffuse scattering from the supermirror coatings will increase the size of the focal spot and influence the signal-to-noise ratio. Investigation on those issues is still underway. With some simple estimations, the neutron intensity can be roughly estimated and is listed in Table 2. It is shown that the intensity for mid-Q configuration is estimated to be over 10 5 n/s, which will still be a significant achievement for SANS at CANS.

Developing competitive SANS instruments at CANS has long been pursued. Among many potential solutions, SANS based on focusing mirrors is a particular promising one. At CPHS, much effort has been made to fabricate nested mirrors with supermirror coatings and a gif-SANS project has been initiated. In this work, we present the preliminary design and the estimated performances of the gif-SANS. The gif-SANS will have three operating modes to cover different Q-ranges and meet different sample requirements. It is shown that such a scheme has great potential to realize a SANS with a high neutron intensity and a wide Q range.