The common characteristics of all histograms are bimodality of the distributions and absence check details of the particles in the range from 40 to 50 nm. The average diameters related to the first part of the size distributions were almost the same for all samples (30 to 35 nm), while in the second part, the average diameter for Si (100) was estimated to be 85 nm; for Si (111), 55 nm; and for both PS samples, 70 to 75 nm. Therefore, in such case, PS sizes of the Cu NPs were not affected by the original Si orientation in contrast to the bulk Si. Such bimodality of the histograms means that the initially deposited Cu

NPs have already coalesced into larger particles (agglomerates) – the second part of the distributions – and new NPs deposited on the reopened surface of the substrates – the first part of the distributions. This mechanism usually takes place in wet depositions [5, 10]. The density of Cu particles on the Si (100) estimated as 109 cm−2 was an order of magnitude

less than those on Si (111) and PS, which are 1010 and 2 × 1010 cm−2 (for the both orientations), respectively. Considering the less density and greater sizes of Cu particles on the bulk Si (100), we suppose that the orientation promotes faster coalescence of Cu buy BMN 673 NPs. Cu NPs have higher mobility due to less number of broken bonds on the Si (100) surface in contrast to Si (111). A greater number of Cu NPs on the PS samples in comparison with bulk Si shows that the porous surface provides more active places for Cu LEE011 supplier adhesion and nucleation. Figure 1 SEM analysis of the surface of samples. (a) Cu/Si (100), (b) Cu/PS/Si (100), (c) Cu/Si (111), and (d) Cu/PS/Si (111). Figure 2 Size distribution histograms. Histograms were made by computer evaluation of SEM images presented on Figure 1. (a) Cu/Si (100), (b) Cu/PS/Si (100), (c) Cu/Si (111), and (d) Cu/PS/Si (111). Microstructure of Cu/Si and Cu/PS/Si samples XRD analysis of the phase composition and crystal orientation of PS after Cu immersion deposition has shown the presence of Cu,

Cu2O, and rarely CuO crystalline phases in the deposit [24]. However, no data dipyridamole were obtained for the initial stages of the Cu immersion deposition because XRD is not sensitive to trace the amounts of crystals of small sizes. To solve the problem, we used EBSD which allows the local study of crystalline object microstructure. Before EBSD analysis, the crystallographic data of the Si, Cu, Cu2O, and CuO phases were entered into the customized HKL channel 5 software database for phase identification. Figure 3 presents the phase maps of the Si and PS surfaces after Cu immersion deposition for 4 s. Table 1 shows the quantitative data of the mapping which resulted in some disagreements with the SEM analysis. According to the phase maps, the Cu amount did not differ greatly for all samples, while the SEM images revealed significant variations of the Cu NP density. We explain it in the following way.