Engineered ion-bombardment as a tool in thin film deposition

Abstract: Energetic-ion bombardment has become an attractive route to modify the crystal growth and deposit high quality thin films. In high-power impulse magnetron sputtering (HiPIMS) where the discharge is characterized by a significant amount of ions of the sputter-ejected atoms, energetic-ion bombardment effects could not be underestimated. This is due to a combination of ions with better mass match to the growing film and reduced incorporation of the working gas. The ions of the sputtered species in HiPIMS also exhibit a broad ion energy distribution function (IEDF), which increases the average energy of the ion flux. However, the IEDF of a HiPIMS discharge is still dominated by ions with energies corresponding to thermalized species. While it is common to use a substrate bias to increase the energy of the incoming ion flux, some deposition set-up make it challenging or even impossible to apply a substrate bias. Hence, it is important to come up with techniques to address such problem. In this thesis, strategies to increase the energy of the ions in a HiPIMS discharge without any substrate bias were investigated and thin films were deposited to evaluate the effects on the crystal growth and film properties.By varying the power per pulse (Ppulse) in a HiPIMS discharge for the deposition of TiN, it was found that the amount of Ti2+ ions could be effectively tuned. The increase in the amount of Ti2+ correlated well with an increase in the film compressive stress, though an optimum Ppulse could be identified in which the stress is relatively low and hardness is high. The results also indicate a broad IEDF for ions of the sputtered species (Ti+, Ti2+, N+) and narrow IEDF for ions of the working gas (Ar+, Ar2+, N2+). However, using such strategy results to an increase in the average ion energy only within the range of <25 eV.Operating a HiPIMS discharge in bipolar mode, where a positive pulse voltage (Urev) is immediately applied after the main HiPIMS pulse, results to ion acceleration due to an increase in the plasma potential to values close to Urev. A simple model is proposed indicating that ion acceleration occurs at a region in between the target and the grounded substrate. Further IEDF measurements with bipolar HiPIMS with varying length of the main HiPIMS pulse (τneg) and length of the positive pulse (τpos) indicate that most of the acceleration happens at the sheath in front of a grounded substrate. By varying τneg and τpos, it was shown that the fraction of accelerated ions can be optimized such that almost all ions are accelerated when τneg is relatively short and τpos is sufficiently long. The growth of TiN films indicate an increase in the compressive stress, hardness and density with an increase in Urev, typical for films deposited using energetic-ion bombardment. Finally, Cu depositions indicate that, due to the pulsed nature of Urev and since most of the ion acceleration happens at the substrate sheath, bipolar HiPIMS process can be compared to a conventional HiPIMS pulse with a synchronized pulsed bias, which has the same timing as the applied positive pulse.

  This dissertation MIGHT be available in PDF-format. Check this page to see if it is available for download.