1. Energy of incident ions

2. Target materials

3. Incident angles of ions

4. Crystal structure of the target surface.

1. Weight loss of target

2. Decrease of target thickness

3. Collection of the sputtered materials

4. Detection of sputtered particles in flight.

2.1.1 Ion Energy

Figure 2.1 shows a typical variation of the sputtering yield with incident ion energy. The figure suggests:

Hull first observed the existence of the sputtering threshold in 1923. He found that the Th–W thermionic cathode in gas rectifier tubes was damaged by bombardment with ions when the bombarding ion energy exceeded a critical value, which was in the order of 20–30 eV.4 The sputtering threshold has been studied by many workers because it is probably related to the mechanism of sputtering. Threshold values obtained by these workers ranged from 50 to 300 eV.5,6 Their results were somewhat doubtful because the threshold energy was mainly determined by measurements of small weight loss from the cathode in the range of 10−4 atoms/ion. The threshold energy is very sensitive to contamination of the cathode surface. In addition, the incident angle of ions and the crystal orientation of cathode materials also change the threshold values.

In 1962, Stuart and Wehner7 skillfully measured reliable threshold values by the spectroscopic method. They observed that threshold values are in the order 15–30 eV, similar to the observations of Hull, and are roughly 4 times the heat of sublimation of cathode materials.

A gas discharge tube used by Stuart and Wehner7 is shown in Fig. 2.2. The target is immersed in plasma generated by a low-pressure mercury discharge. In a noble gas discharge, the mercury background pressure is about 10−5 Torr (1.33×10−3 Pa) or less. The temperature of the target is kept at about 300°C so as to reduce condensation of the mercury vapor. The sputtered target atoms are excited in the plasma and emit the specific spectrum. The sputter yields at low energy are determined by the intensity of the spectral line. This technique eliminates the need for very sensitive weight measurements.

Typical sputter yields in a low-energy region measured by Stuart and Wehner are shown in Fig. 2.3. The threshold values determined in the sputter yields are in the order of 10−4–10−5 atoms/ion. Table 2.1 summarizes the sputtering threshold energy measured by the spectroscopic method for various target materials.7,8

Table 2.1 Sputtering Threshold Data7,8

H, heat of sublimation.

The table suggests that there is not much difference in threshold values. The lowest value, which is nearly equal to 4 times the heat of sublimation, is observed for the best mass fit between target atom and incident ion. The higher threshold energy is observed for poor mass fits.

The threshold energy also strongly depends on the particular sputtering collision sequence involved. High threshold energy (i.e., Eth/U0>10 for Ar+→Cu, where Eth denotes the threshold energy; U0, the heat of sublimation) will be expected in the collision sequence where primary recoil produced in the first collision is ejected directly. Lower values will be observed for the multiple sputtering collisions. An incident angle of around 40–60° offers the minimum threshold energy (i.e., Eth/U02 for Ar+→Cu) under the multiple sputtering collisions.9

The sputter yield varies with the incident ion energy E. In the low-energy region near the threshold, S obeys the relation S∝E2 as seen in Fig. 2.1. This occurs at the energy region in the order of 100 eV, S∝E.6,10 In this energy region, the incident ions collide with the surface atoms of the target, and the number of displaced atoms due to the collision will be proportional to the incident energy.

At higher ion energies of 10–100 keV, the incident ions travel beneath the surface and the sputter yields are not governed by the surface scattering but by the scattering inside of the target. Above 10 keV, the sputter yields will decrease due to energy dissipation of the incident ions deep in the target.

Maximum sputter yields are seen in the ion energy region of about 10 keV. Figure 2.4 summarizes the energy dependence of the yield, i.e., the sputter rate, as reported by Sigmund.11

2.1.2 Incident Ions, Target Materials

Sputter yield data have been extensively accumulated in relation to gas discharges, sputter deposition, etching, surface analysis, and radiation damage. At first, sputter yields were measured in the cold cathode discharge tube.12,13 But these measurements did not offer reliable data because the incident ions and sputtered atoms frequently collide with discharge gas molecules in the cold cathode discharge tube.

Laegreid and Wehner,14 in 1959, accumulated the first reliable data of sputter yields in a low gas pressure discharge tube. At present, this yield data is still widely used for sputtering applications. Figure 2.5...