Experimental data for "Superradiant Thomson scattering from graphite in the extreme ultraviolet"

Pretreated experimental data of an extreme ultraviolet (EUV) Thomson scattering experiment on highly oriented pyrolytic graphite (HOPG), described in the PNAS paper "Superradiant Thomson scattering from graphite in the extreme ultraviolet" by the same authors. Details on the experimental geometry and setup are available in the paper. 

The data contain information on the distribution of the Thomson scattering intensity, at varying the incident intensity. Given the observation that the ratio of the scattered to incident intensity is not constant throughout the experiment, but depends on the incident intensity, the pulse-by-pulse measured intensities are treated as follows. 

Each independent dataset (at given incident wavelength and polarization) of our experiment was treated by averaging the ratio of I_scat to the incoming I_0, R_scat (I_0) = I_scat (I_0)/I_0, over several different sample points. The average was taken in small enough intervals of I_0. Statistically meaningful data were obtained rejecting I_0 intervals with too few (<5) data within them and (rare) fluctuations larger than twice the standard deviations. These data are reported in the available files. 

The files are named after the value of the wavelength and polarization of the EUV free electron laser beam, exciting the HOPG sample.

"lambda0-4nm" -> incident wavelength 4.08 nm"lambda0-5nm" -> incident wavelength 4.74 nm"pol-s" -> polarization orthogonal to the scattering plane"pol-p" -> polarization parallel to the scattering plane

The file column are as follows: col. 1: incident EUV pulse intensity I_0 (central values for of the sampled intervals, see above)col. 2: ratio between scattered and incident intensity R_scat = I_scat/I_0col. 3: experimental error associated to col. 2, estimated as the standard deviation in the data relative to a given I_0 intervalcol. 4: scattering angle.

Col. 1,2,3 are in arbitrary units, as discussed in detail in the paper. 

Please note that in the case of lambda0-4nm, pol-s, two datasets are available, relative to two different graphite samples: these differ for a scaling value of I_0 (about the 20%), possibly associated to a slightly different surface quality and/or focusing conditions.