The transient and steady state thermal flux at the substrate during the deposition of aluminum film in a direct current magnetron sputter system has been determined by measuring the resistance of a complementary metal–oxide–semiconductor (CMOS) sensor. The sensor is calibrated using ohmic self-heating before the plasma is switched on. The steady state thermal flux at the substrate was measured to vary from 9.6 to 46 mW/cm2 at a substrate-target distance of 10.8 cm depending on the magnetron power (75–300 W) and gas pressure. Plasma radiation and electron bombardment are noted to be the most significant sources of the thermal flux to the substrate, each contributing about 36% and 29%, respectively, of the total thermal flux at the substrate for a magnetron power of 200 W and gas pressure of 5 mTorr. Thermal radiation is also an important factor, along with kinetic energy and condensation energy. Total energy per deposited atom is calculated to be in the range of 28–52 eV depending on the magnetron power and gas pressure, and increases with pressure but decreases with magnetron power. The trend seems to suggest that at higher magnetron powers (>300 W for a 3 in. target), a pressure independent total energy per deposited atom may be obtained.

The energy flux onto an unbiased substrate is determined theoretically and experimentally for aluminum and copper deposited using a 3 in. magnetron sputtering system. The energy per deposited atom is calculated. Energy per deposited atom trends towards being independent of power and pressure, especially at high magnetron powers. At low powers, the energy per deposited atom increases with pressure due to lower deposition rates. For the magnetron system used, plasma effects are shown to be important in determining the total energy flux to the substrate. Contributions of the electrons and thermal radiation from the target region are included in the model.

No abstract

3D numerical simulation of gas heating in a magnetron sputtering system is performed. Pressure, magnetron power density and location of the substrate plane in front of the target are shown to affect the gas temperature profile. For the pressure range under study, maximum gas temperature is shown to increase with pressure. By increasing the separation between the target and the substrate, the maximum gas temperature is shown to increase up to the point when most of the particles are assumed thermalized. Cu shows more gas heating than does Al due to its higher sputtering yield and energy transfer efficiency. Sputtered particles are shown to be the major source of gas heating.

Variations in the magnetic field strongly affect the plasma parameters in a magnetron sputtering system. This in turn affects the throughput as well as the energy flux to the substrate. The variation in the magnetic field in this study, for a dc magnetron process, is achieved by shifting the magnet assembly slightly away from the target. Measurements of the plasma parameters show that while the electron density at the substrate increases with decrease in magnetic field, the electron temperature decreases. The cooling of the electron temperature is consistent with results reported elsewhere. The deposition rate per input magnetron power is found to increase slightly with the decrease in magnetic field for the process conditions considered in this study. Results suggest that the energy flux to the substrate tends to show a general decrease with the shift in the magnet assembly.