The main requirements for a reflectometer are high flux, flexible resolution, accurate collimation and precision neutron detection. D17 has all of these properties. This article shows some of the first results taken with D17 using the time-of-flight (TOF) option including data from the first official experiment.

The principal reason for having a TOF option on this instrument was to enable experiments on the kinetics of planar materials to be performed. This is possible because a range of q is measured simultaneously using a white beam as opposed to the monochromatic mode where the sample angle has to be rotated for each point in q.  The TOF is less efficient than the monochromatic mode in terms of the flux available at each q point as the flux at the extremes of usable wavelengths are much lower than the peak flux used in the monochromatic mode. Having a chopper system where the resolution can be continuously varied, however, compensates for this as flux can be gained when high resolution is not required. A second advantage is that when setting up experiments it is easy to check the alignment of the sample after only a few seconds of data acquisition. For example figure 1 shows raw data from counting for only 10 seconds. The vertical resolution of the detector has been summed resulting in a picture with X-pixels horizontally, corresponding to 2q, and the vertical pixels representing wavelength. It can be seen that above the critical wavelength where total reflection occurs there is no intensity in the direction of the incoming beam. This means that no neutrons fly past the sample surface and this surface is under illuminated and correctly aligned with respect to the incoming beam. The picture demonstrates the fact that neutrons obey simple optical laws.  It can be seen that below the critical wavelength the transmitted beam is refracted towards the horizon showing that neutrons, unlike light, have a refractive index less than one (for a material with positive scattering length density). The curvature of the diffracted beam with wavelength is due to the fact that the formula for neutron refractive index has a wavelength squared term.

For the first officially scheduled experiment (July 2000) neutron reflectivity curves were measured from adsorbed lipid bi-layers at a silicon/water interface[1,2]. The lipid layers undergo a phase transition from a gel to a fluid phase and it is known that around the transition a substantial increase of the thickness at the interface is due to an increase in the quantity of water between the two bi-layers. The q-range explored on D17 (see Figure 2) was larger than that measured in the past on this kind of sample as the lowest reflectivity is usually limited by background from bulk water and sample cell. With the high flux and wide-angle multi-detector reflectivities of ~10-7 could be measured in the reasonable time of one hour. Analysis is still in progress but the increased q-range should give additional information for modelling cell membranes. As the water thickness is largely defined by the first minimum in the reflectivity, a limited q -range (0.005-0.05 Å-1 measurable in 7 minutes) is sufficient to follow the water thickness through the transition. In the top-right insert of figure 2 the fitted thickness of the water layer between the two bi-layers is given at the different temperatures shown.
In addition to the ability to subtract incoherent backgrounds more efficiently there is another advantage to having a large area multi-detector. About the specular reflection there can be off-specular scattering arising from surface roughness or as in the case described below, low angle diffraction. The sample was a diffraction grating consisting of a glass substrate with thousands of strips of nickel, 900Å deep separated by 10mm       (Ott and Menelle, Saclay). If this grating were placed perpendicular to the incoming beam then 10Å neutrons would be diffracted only a few thousandths of a degree (qx=6x10-5   Å-1), beyond the range of any existing small angle instrument. However, if the incoming beam strikes the grating surface at a glancing angle then many diffraction orders can be seen in both the reflected and transmitted beams at measurable angles of deflection (figure 3). In addition to information on the stripe separation, the depth profile (consistent with 900Å) is revealed in the ripples of intensity found running along the specular line and the various reflected diffraction orders. The diagonal line of intensity coming from where the specular line just totally reflects is a Yoneda wing and is a consequence of the roughness along the surface of the nickel strips. D17 is ideally suited  to investigate not just structures as a function of depth but also within the plane such as magnetic grain boundaries or polymer droplets.

Recently both the monochromatic and the polarised neutron modes of the instrument have been successfully tested. The polarisation analyser will be installed in the winter shutdown 2000.

Useful wavelength range 2-20Å
Qz range (normal to a surface) 2x10-3 - 3 Å-1
Qx range (within a surface): 1x10-6 - 2x10-2 Å-1
white beam flux   1010 n/s/cm2
Detector size 250 x 500 mm
Detector resolution 1.5 x 3 mm

Refs:

[1] Charitat et al., Eur. Phys. J. B, 1999
[2] Fragneto et al., ILL Annual report 1999
 

Figures