The primary objective of this study was to investigate and compare the surface topography of hard coatings deposited by three different physical vapor deposition methods (PVD): low-voltage electron beam evaporation, unbalanced magnetron sputtering and cathodic arc evaporation. In these deposition systems, various ion etching techniques were applied for substrate cleaning. The paper summarizes our experience and the expertise gained during many years of development of PVD hard coatings for the protection of tools and machine components. Surface topography was investigated using scanning electron microscopy (SEM), atomic force microscopy (AFM), scanning transmission electron microscopy (STEM) and 3D stylus profilometry. Observed similarities and differences among samples deposited by various deposition methods are discussed and correlated with substrate material selection, substrate pretreatment and deposition conditions. Large variations in the surface topography were observed between selected deposition techniques, both after ion etching and deposition processes. The main features and implications of surface cleaning by ion etching are discussed and the physical phenomena involved in this process are reviewed. During a given deposition run as well as from one run to another, a large spatial variation of etching rates was observed due to the difference in substrate geometry and batching configurations. Variations related to the specific substrate rotation (i.e., temporal variations in the etching and deposition) were also observed. The etching efficiency can be explained by the influence of different process parameters, such as substrate-to-source orientation and distance, shadowing and electric field effects. The surface roughness of PVD coatings mainly originates from growth defects (droplets, nodular defects, pinholes, craters, etc.). We briefly describe the causes of their formation.

Total sputtering yield and spatial distributions of sputtered atoms are important for numerous deposition techniques. We performed SRIM (Stopping and Range of Ions in Matter) simulations to analyze the total sputtering yield and angular distribution of sputtered atoms for a range of single-element target materials. The simulations were conducted for normal argon ion incidence in the 300–1200 eV range and at an oblique angle for selected ion energies. We examined the total and differential sputtering yields for the transition metals in the periods 4–6 and groups 4–6 (Ti, V, Cr; Zr, Nb, Mo; Hf, Ta, and W) and group 11 (Cu, Ag, and Au) of the periodic table, and other materials that are relevant to sputtering (B and C; Al and Si). For the transition metals, the total sputtering yield increases with the group of the periodic table. The elements in group 4 (i.e., Ti, Zr, and Hf) have the lowest sputtering yield, while the elements in group 11 (i.e., Cu, Ag, and Au) exhibit the highest sputtering yield. The angular distribution of the sputtered atoms shows a cosine distribution for the transition metal atoms. The angular distribution of the sputtered atoms for the oblique ion incidence is more asymmetric for the lower ion energies, while for the higher ion energies, the atoms are sputtered more symmetrically. The symmetry also depends on the group of the periodic table and the atomic mass of the target material. The elements in group 11 show the most symmetric distribution, while the elements in group 4 experience the most asymmetric distribution. Furthermore, in an individual group, the distribution becomes more symmetric with heavier target elements. We also examined in detail the influence of the surface binding energy, atomic mass, and ion energy on the total sputtering yield. These parameters were analyzed with regard to the simplified analytical formula for the total sputtering yield, which was derived by Sigmund. This formula was modified by introducing a power fitting parameter, which accounts for the non-linear sputtering yield dependence on the ion energy. The equation provided good estimates for the total sputtering yield of the transition metals that were sputtered by argon ions with energies up to 1200 eV.

The energy of the sputtered atoms is important to control the microstructure and physical properties of thin films. In this work, we used the SRIM program to simulate the energy of sputtered atoms. We analyzed the energy distribution functions (EDFs) and the average energies of the atoms in different spatial directions for a range of target materials and Ar ion energies. The results were compared to the analytical equations for EDFs derived by Sigmund and Thompson and with experimental data from the literature. The SRIM simulations give realistic EDFs for transition metals, but not for elements lighter than Si. All EDFs show a low-energy peak positioned close to one-half of the surface binding energy and a high-energy tail decreasing as approximately E⁻². We analyzed the characteristics of EDFs, specifically, the position of low- and high-energy peaks, FWHM, and the energy tail, with respect to the ion energy and position of the element in the periodic table. The low-energy peak increases with atomic number for elements within each group in the periodic table. Similar changes were observed for FWHM. For the period 5 and 6 elements, additional broad high-energy peaks were observed at emission angles above 45° when sputtered by Ar ions with 300 eV and also in some heavier elements when bombarded by 600 eV and 1200 eV ions. The transition metals in groups 4, 5, and 6 in periods 5 and 6 have the highest average energies, while the lowest average energies have elements in group 11. The results of simulations show that the average energies of sputtered atoms were inversely proportional to the sputtering yield, i.e., the higher the sputtering yield, the lower the average energy of sputtered atoms. We established an empirical equation for transition metals to estimate the average energy of sputtered atoms from the sputtering yield. The angular distribution of the average atom energy depends on the atomic number. Transition metals with 22 < Z < 72 have an anisotropic energy distribution, with the highest average energies in the 40°–70° range. For the elements in group 11, the angular distribution of the average energies is more isotropic.

The authors wish to make the following correction to this paper