Specular reflectometry simulation

To set up and run a specular reflectometry simulation, use

import bornagain as ba
scan = ...
sample = ...
simulation = ba.SpecularSimulation(scan, sample)
# ... set options
result = simulation.simulate()

Supported scan types: grazing angle scan, wavelength scan, wavenumber qz scan.

For the constructor arguments, see sections scan and sample.

For optional settings, see simulation options.

For the return type of function simulate(), see Datafield.

Output intensity

simulate() returns reflected intensity, with reflectivity R:

I(x) = apply_background_to(I0 * R(x) * footprint(x))

Set I0 with scan.setIntensity(I0); the default is one.

If the medium of incidence absorbs, incident and reflected amplitudes refer to the lower interface of the top layer, or to the upper interface of the bottom layer for a beam from below. Attenuation between that plane and the physical source or detector is not included.

Applied magnetic field

If the sample is in an external field, the field also permeates the top layer, where the two spin states of the neutron have different kinetic energies (Zeeman splitting). BornAgain then treats the two spin eigenmodes along the field as separate channels:

  • The glancing angle α refers to the beam in the top layer, with the magnetic contribution to its potential omitted: for a nonabsorbing top layer with nuclear refractive index n, the vertical wavenumber of the incident spin state is n (2π/λ) sin α, with λ the vacuum wavelength. For a beam from vacuum or air, α is the usual glancing angle. On entering the field, this vertical wavenumber is conserved.
  • A spin-flipped reflected neutron has a different vertical wavenumber, and leaves at a different angle. Its reflectivity, reported at the incident angle, is the ratio of the outgoing to the incident vertical flux: the squared amplitude times the ratio of the vertical wavenumbers. Below a cutoff angle, the flipped state of one channel cannot propagate in the top layer, and that spin-flip reflectivity is zero.
  • Polarizer and analyzer must be parallel or antiparallel to the field, or absent. The model does not compute the phase that the two spin states accumulate along the beam path through the field outside the sample. For these orientations, the intensity does not depend on that phase: it is the sum of the channel reflectivities, weighted with the polarizer populations and the analyzer transmissions of the two states. Imperfect polarization and analyzer efficiencies are allowed. Other orientations are rejected.
  • Without a field in the top layer, the spin states of a beam from above are degenerate, the channels are summed coherently, and polarizer and analyzer may point in any direction.

This follows Maranville et al., J. Appl. Cryst. 49, 1121 (2016), and the reference code Refl1D, except that Refl1D reports squared amplitudes without the flux factor.

For a beam from below, the bottom layer is the medium of incidence; a magnetic induction in the bottom layer is supported only if all inductions in the sample are parallel or antiparallel to it, and polarizer and analyzer must then be parallel or antiparallel to it as well.

Example: Zeeman splitting.

Examples
Further reading