Sputtering, physical

Measured sputtering yields for nickel by H , Ne , and Ni+ ion bombardment versus incident energy are displayed in Fig. 13. Also shown are theoretical calculations based on TRIDYN code (Biersack and Eckstein, 1984) and on the Bohdansky (1984) formula. It is rather obvious that the sputtering yields for the heavier Ne and Ni projectiles are quite different, both in magnitude and with respect to the position of the maximum, from those with H impact (after Eckstein, 1991). 

Sputtering yield of nickel for and Ne ion impact vs. incident energy (after Eckstein, 1991). 

The following empirical expressions for / and rj have been given by Yamamura etal. (1983, 1984)  

Within the collision cascade model, the energy distribution N E) of sputtered atoms is proportional to 

Erosion due to energetic particle bombardment depends on a number of parameters such as mass ratio of incident particles to surface atoms, particle energy and flux, as well as surface temperature. In the following, the physical understanding of different erosion mechanisms occurring at different wall components in fusion devices will be presented together with supporting data from laboratory experiments. 

---

Barone M E and Graves D B 1995 Chemical and physical sputtering of fluorinated silicon J. Appi. Phys. 77 1263-74... 

R. Behrisch, ed., "Sputtering by Particle Bombardment I Physical Sputtering of Single Element SoHds," in Topics inMppliedPhjsics, Vol. 47, Springer-Vedag, Berlin, 1981. 

W. Eckstein and V. Philipps, Physical Sputtering and Radiation-Enhanced Sublimation. In Physical Processes of the Interaction of Fusion Plasmas with Solids, W.O. Hofer and J. Roth, Eds., 1996, Academic Press, pp. 93-134. 

Figure 1(a) shows the etch rates of niobium oxide pillar and Si film, and the etch selectivity of Si to niobium oxide as a function of CI2 concentration. The etch condition was fixed at coil rf power of 500 W, dc-bias to substrate to 300 V and gas pressure of 5 mTorr. As the CI2 concentration increased, the etch rate of niobium oxide pillar gradually decreased while Si etch rate increased. It indicates that the etch mechanism of niobium oxide in Cl2/Ar gas is mainly physical sputtering. As a result, the etch selectivity of Si film to niobium oxide monotonously increased. The effect of coil rf power on the etch rate and etch selectivity was examined as shown in Fig. 1(b). As the coil rf power increased, the etch rates of niobium oxide and Si increased but the etch rate of niobium oxide showed greater increase than that of Si. It is attributed to the increase of ion density with increasing coil rf power. Figure 1 (c)... 

It is important to note that Eqs. 5, 8, and 9 were derived entirely from a silicon material balance and the assumption that physical sputtering is the only silicon loss mechanism thus these equations are independent of the kinetic assumptions incorporated into Eqs. 1, 2, and 7. This is an important point because several of these kinetic assumptions are questionable for example, Eq. 2 assumes a radical dominated mechanism for X= 0, but bombardment-induced processes may dominate for small oxide thickness. Moreover, ballistic transport is not included in Eq. 1, but this may be the dominant transport mechanism through the first 40 A of oxide. Finally, the first 40 A of oxide may be annealed by the bombarding ions, so the diffusion coefficient may not be a constant throughout the oxide layer. In spite of these objections, Eq. 2 is a three parameter kinetic model (k, Cs, and D), and it should not be rejected until clear experimental evidence shows that a more complex kinetic scheme is required. 

---