Neutron phase filtering for separating phase- and attenuation signal in aluminium and anodic aluminium oxide

TL;DR

This paper compares neutron phase filtering methods for separating phase and attenuation signals in aluminium and anodic aluminium oxide, highlighting the advantages of a simulation-based approach for improved interface analysis.

Contribution

It introduces a simulation-based assessment using ray-tracing software McStas to improve phase filtering in neutron imaging, surpassing traditional transport-of-intensity methods.

Findings

Simulation-based approach outperforms phase filter method

McStas software enables better separation of phase and attenuation signals

Potential for quantitative analysis at refracting interfaces

Abstract

Neutron imaging has gained significant importance as a material characterisation technique and is particularly useful to visualise hydrogenous materials in objects opaque to other radiations. Particular fields of application include investigations of hydrogen in metals as well as metal corrosion, thanks to the fact that neutrons can penetrate metals better than e.g. X-rays and are at the same time highly sensitive to hydrogen. However at interfaces for example those that are prone to corrosion, refraction effects sometimes obscure the attenuation image, which is used to for hydrogen quantification. Refraction, as a differential phase effect, diverts the neutron beam away from the interface in the image which leads to intensity gain and intensity loss regions, which are superimposed to the attenuation image, thus obscuring the interface region and hindering quantitative analyses of e.g. hydrogen content in the vicinity of the interface or in an oxide layer. For corresponding effects in X-ray imaging, a phase filter approach was developed and is generally based on transport-of-intensity considerations. Here, we compare such an approach, that has been adapted to neutrons, with another simulation-based assessment using the ray-tracing software McStas. The latter appears superior and promising for future extensions which enable fitting forward models via simulations in order to separate phase and attenuation effects and thus pave the way for overcoming quantitative limitations at refracting interfaces.