A design for an incident-beam collimator for the Paris–Edinburgh pressure cell is described here. This design can be fabricated from reaction-bonded B4C but also through fast turnaround, inexpensive 3D-printing. 3D-printing thereby also offers the opportunity of composite collimators whereby the tip closest to the sample can exhibit even better neutronic characteristics. Here, we characterize four such collimators: one from reaction-bonded B4C, one 3D-printed and fully infiltrated with cyanoacrylate, a glue, one with a glue-free tip, and one with a tip made from enriched 10B4C. The collimators are evaluated on the Spallation Neutrons and Pressure Diffractometer of the Spallation Neutron Source and the Wide-Angle Neutron Diffractometer at the High Flux Isotope Reactor, both at Oak Ridge National Laboratory. This work clearly shows that 3D-printed collimators perform well and also that composite collimators improve performance even further. Beyond use in the Paris–Edinburgh cell, these findings also open new avenues for collimator designs as clearly more complex shapes are possible through 3D printing. An example of such is shown here with a collimator made for single-crystal samples measured inside a diamond anvil cell. These developments are expected to be highly advantageous for future experimentation in high pressure and other extreme environments and even for the design and deployment of new neutron scattering instruments.
REFERENCES
First, the experimentalsetup on WAND2 used an intermediate design of the radial oscillating collimator with very broad collimator blades. In addition, the shielding was not completed so that a large uniform background was detected. Since intensities for the vanadium standard were not comparable to the measured data signal, the vanadium calibration led to an under-correction. Thus, the detector channel structure remained visible, especially in places of high background, i.e. at low Q. The second artifact stems from a grounding issue that produced additional electronic noise in the first segment of the detector, from 3° to 18°. This artifact led to a reduced efficiency. In this case, the different count rates for vanadium standard and sample led to an over-correction for all channels. This manifests in a step-like transition between the segments seen in the data.
