During the irradiation, the base pressure of chamber was maintained at approximately 10−7 mbar. The ion beam current density
was kept constant at 15 μA/cm2. The beam was scanned uniformly over an area of 10 mm × 10 mm by electromagnetic beam scanner. After irradiation, the samples were analyzed by Nano Scope IIIa atomic force microscope (AFM; Bruker AXS Inc, Madison, WI, USA) under ambient conditions in tapping mode. Cross-sectional transmission electron microscopy (XTEM) was carried using a Tecnai-G2-20 TEM (FEI, Hillsboro, OR, USA) facility operating at 200 kV. The cross-sectional specimens for TEM study were prepared by Ar ion beam milling at 4 kV/20 μA and at an angle of 4° with respect to the sample surface. Figure 1 Schematic view
of 50 keV Ar + ion beam irradiation. For first stage (to prepare two deferent depth locations of a/c interface) at an angle of (a) 60° and (b) 0°, CX-5461 mw with respect to surface normal; second stage irradiation (for fabrication of ripples) at an angle of 60° named as (c) set A and (d) set B. Testing the hypothesis AFM characterization was carried out on all samples after each irradiation step. After first irradiation, the average RMS roughness for both sets of the samples was nearly similar Raf inhibitor (0.5 ± 0.1 and 0.6 ± 0.1 nm). In the second stage, all samples were irradiated by a stable 50 keV Ar+ at same angle of incidence (60°) for all fluences. Figure 2a,b,c,d, and e,f,g,h shows the AFM images for set A and set B samples after the second stage irradiation at the fluences of 3 × 1017, 5 × 1017, 7 × 1017, and 9 × 1017 ions per square centimeter, respectively. It was found that for set A, the wavelength and amplitude were increasing with increase in irradiation fluence (as shown in Figure 3). For set B samples, the average wavelengths of ripples were nearly same cAMP as that of set A samples at corresponding fluences. However, the observed average amplitudes of ripples are about one order less in magnitude for set B as compared to those for set A since the only difference between two sets of samples was
the depth location of a/c interfaces. If the evolution ripples were based on curvature-dependent sputtering and surface diffusion, we should have got ripples of identical dimensions for corresponding equal fluence in both sets of samples. Despite similar initial surface morphology of both sets of samples after first stage of irradiation, the observation of similar wavelength and lower amplitude of ripples in set B samples as compared to set A samples casts doubt on the validity of Bradley-Harper and its extended theories. It can be emphasized here that we repeated complete set of experiment with two different ion beams and at different energies (Ar at 50 keV and Kr at 60 keV). And in all cases, the observed trend was similar. To the best of the authors’ knowledge, there is no existing model which could physically explain this anomaly.