ROG, the neutron reflectometer at IRI, Delft

Towards a detailed resolution smearing kernel for time-of-flight neutron reflectometers

“…However, if the collimation system is some distance after the chopper discs, the beam height profile in the chopper system (that would be accepted by the collimation systems) is divergent and would be similar to Fig. 5 of de Haan et al (1995), giving the distribution for p h ðÞ shown in Fig. 4.…”

“…Unfortunately, NR suffers the drawback that the source (whether a reactor or spallation source) brilliance is many orders of magnitude lower than most X-ray sources, even the lowly Cu K tube in a laboratory. Most neutron reflectometers can be divided into two classes, those that are angular dispersive (Dura et al, 2006) and those that are energy dispersive (Campbell et al, 2011;Webster et al, 2006;James et al, 2011;de Haan et al, 1995). The former use a monochromator to select a specific wavelength band to direct onto the sample -NR measurements are performed at many different angles of incidence.…”

“…Energy dispersive measurements have the advantage of being able to measure across a broad range of momentum transfer at a single setting, making them attractive for those who want to carry out kinetic measurements. Energy dispersive reflectometers are found at spallation sources (Webster et al, 2006), and at reactors (Campbell et al, 2011;James et al, 2011;de Haan et al, 1995;Cousin et al, 2011), where the timeof-flight setup is based on the use of choppers to pulse the beam.…”

“…To make most efficient use of the available flux, NR measurements are typically performed at a more relaxed Q z resolution than XRR measurements. There are two main components that contribute to resolution: (1) the angular divergence of the collimated beam, and (2) the wavelength resolution of the radiation (de Haan et al, 1995;van Well & Fredrikze, 2005). It is essential that these resolution terms be accounted for when analysing reflectivity data, because they smear sharp features present in the curves.…”

“…The functional forms of both the angular and wavelength components are understood and have been explained in detail in articles by de Haan et al (1995) and van Well & Fredrikze (2005). However, van Well and Fredrikze assume that the overall resolution kernel is Gaussian in shape, with the width of this Gaussian being formed by adding the fractional variances of all the components [equation 3], where there is no correlation between wavelength and angle of incidence:…”