Figure 1

(a) Young’s interference fringes used to measure the coherence length of the x rays at the plane of sample. The data are analyzed using a Gaussian statistically stationary model for the coherence function. The exit slit of the monochromator had a nominal width of $525\mu \mathrm{m}$ for these data. (b) Scanning electron micrograph of the sample used in the CDI experiments. The sample is drilled in gold using a focused ion beam. The sample dimensions are shown. (c) Diffraction data from the sample in (a) obtained using 1.4 keV x rays. This data was obtained using the highest coherence available. The color bar along the bottom of the image indicated the relative intensity of the measured data, with the highest intensity normalized to unity and corresponds to a scattering vector at the detector corners with magnitude $9.8{\mathrm{nm}}^{-1}$. (d) Detail from the data. The panel shows the region of the data contained in the white box seen in (b) for highly coherent data to the left of the red line and low coherence data to the right.Reuse & Permissions

Figure 2

(a) Reconstruction of the magnitude of the wave leaving the sample using the known coherence information (labeled “multimodal” propagation). (b) A reconstruction of the same data in identical conditions except that full spatial coherence is assumed (labeled “coherent” propagation). We were not able to obtain a reconstruction from this data under this assumption. (c) Reconstruction of the high-spatial coherence ($5\mu \mathrm{m}$ exit slit setting) data using the algorithm that allows for a deviation from perfect coherence. (d) Reconstruction using an algorithm that assumes perfect spatial coherence.Reuse & Permissions