Np. Barradas et R. Smith, Simulated annealing analysis of nuclear reaction analysis measurements of polystyrene systems, J PHYS D, 32(22), 1999, pp. 2964-2971
Nuclear reaction analysis (NRA) is a standard technique in compositional an
alysis of materials, including depth profiling of elements. It is particula
rly useful for the profiling of deuterium in polymers. However, the data an
alysis can be cumbersome and time-consuming. In this paper we present fully
automated analysis of NRA data using the simulated annealing combinatorial
optimization algorithm. The routine is general and can be used to analyse
data from any reaction with known Q value and cross section. The shape of t
he cross section is taken into account, and hence depth profiling using rea
ctions with non-hat cross sections can be done in a trivial way. Furthermor
e, the algorithm allows one to take into account the degradation of energy
resolution with probing depth, which is shown to be crucial in analysing th
in layers, where the depth resolution is a determining factor. We have also
shown that we can calculate confidence limits on the depth profiles obtain
ed by using Bayesian inference with the Markov chain Monte Carlo method. We
present the application of the algorithm to the analysis of deuterated pol
ystyrene samples.