Arvydas, Asger, and Kasper

V-XX-MMXIII

Neutron Reflectometry
Neutron reflectometry is a specular reflection technique used for measuring the structure of thin
films, including the thickness, density, and roughness of any thin films layered on the substrate. It is
similar to x-ray reflectivity and ellipsometry, where the angle of the
incident beam is equal to the angle of the reflected beam. An
advantage of neutron reflectometry is that is based on nuclear
contrast rather than electron density, which makes it more sensitive
when measuring, lighter elements such as H, C, N, O, etc. Since
neutrons can penetrate deep into matter and typically non-perturbing
it allows the use of delicate sample materials, such as biological membranes. The disadvantages are
mainly the more costly infrastructure needed, and that some materials may become radioactive.
When compared with x-ray reflectivity resolution may be more limited.
Method
In order to do neutron reflectometry you have to know some of the properties of the constituents of
the sample, such as molecular mass, and density. When working with lipid compounds most of the
information can be obtained from Avanti Polar Lipids, Inc. (avantilipids.com). For the simulation of
the experiment we also had to look up the molecular structure and volume. Inputting these values
for each of the possible constituents in Motofit (a plug-in for Igor) we get the corresponding neutron
scattering length density (neutron SLD), from these values we are able to simulate any composition
of the constituents. One of the effects we can illustrate with the simulations is the importance of the
correct choice of background material as it must give a good contrast for the sample. In our case we
compared samples in water (H 2 O) with the same samples in heavy water (D 2 O). The best contrast is
obtained when there is the greatest difference between the SLD of the sample material and the
background. If the difference in SLD is too small the sample cannot be distinguished from the
background. Furthermore by adding an additional layer it is possible to generate interference
patterns called fringes, while measuring thin layers that might otherwise not exhibit these patterns.
This makes it possible to calculate the thickness of the sample, based on the interim between each
local minima (or maxima) as illustrated.
=

2
,

where is the thickness of the layer and is the momentum

transfer vector.

Arvydas, Asger, and Kasper

V-XX-MMXIII

The Simulation
The initially gathered data for the simulation can be seen in table 1.

H2O
D2O
PS
d8-PS
DPPC
Si
SiO2

Water
Heavy water
Polistyrene ((C8H8))
STYRENE-D8
Dipalmitoylphosphatidylcholine
Silicon
Silicon dioxide

Molec mass g/mol

18.0153
20.0276
104.1498
336.6000
734.0390
28.0850
60.0800

Density g/cm^3
1.000
1.107
1.000
1.000
1.050
2.329
2.648

Table 1: The different constituents used during simulation and their inherent properties.

These values were used to simulate samples consisting of different lipid layers (one or two) on
silicon wafers in either water or heavy water.
Below are some simple simulations of water and heavy water. The materials are arranged so that the
incident beam first hits the last mentioned material. In the case of Heavy water on Si Wafer, the
incident beam first hits the Si Wafer.
Notice total reflection occur in two of the cases (R=1). This phenomenon happens when an incident
beam hits a material with a lower refractive index at or above the critical angle. This implies the
following order of refractive indices when dealing with neutron reflectometry: 2 < < 2 .

Below we have three examples of the impact that the choice of background has on the results:

Arvydas, Asger, and Kasper

V-XX-MMXIII

Arvydas, Asger, and Kasper

V-XX-MMXIII

The thicknesses of the single layers are calculated in table 2. Multilayered samples are a more
complicated matter.
Sample
Thickness of layer

H 2 O-DPPC-SiO 2 -Si
31.73

D 2 O-DPPC-SiO 2 -Si
30.51

H 2 O-PS-SiO 2 -Si
389.71

Table 2: Calculated values of samples thicknesses

D 2 O-PS-SiO 2 -Si
391.23

When multiple maxima/minima are identifiable in a reflectometry graph, it is preferable to take an

average of in order to reach a more precise result:
=

max() max(1)
1

In the case of the DPPC-graphs, it is barely possible to find two following minima/maxima, so in such
cases one or more extrema has to be estimated since there might not be an easily distinguishable
extremum.

Conclusion
We have gained some experience simulating neutron scattering of thin soft films. This has given us a
feeling for some of the important factors and capabilities of neutron reflectometry. We have noted
that the refractive indices may depend on the wavelength non-linearly. They may also be very
different for neutrons compared with electromagnetic waves. For instance in the case of Silicon and
heavy water, Silicon has the higher index for electromagnetic waves, whereas this relation is
reversed for neutrons.