Deposition rate and energy to substrate in chopped and standard HiPIMS: identifying optimal pulse parameters

TL;DR

The paper investigates how micropulse length, inter-pulse delay, and magnetic field strength affect energy flux, deposition rate, and ionized flux in chopped versus standard HiPIMS under fixed average and pulse power. Using a combination of a passive thermal probe, QCM ion meter, and in-situ mass spectrometry, the authors reveal that chopped HiPIMS generally increases energy delivery and deposition relative to standard modes, with the strongest gains under certain magnetic-field configurations. However, standard HiPIMS with short pulses and high repetition frequency can outperform chopped HiPIMS under equivalent total pulse length, emphasizing that short pulses and adequate off-times for gas refilling are crucial for maximizing ion flux and deposition. The findings highlight the context-dependent nature of pulse optimization in HiPIMS and provide guidance for tuning pulse architecture to achieve desired deposition characteristics in Ti-based films.

Abstract

High-Power Impulse Magnetron Sputtering (HiPIMS) offers higher ionized flux fractions at the cost of lower deposition rates compared to conventional DCMS. A fine optimization of the deposition conditions is crucial for specific applications. Chopped or multi-pulse HiPIMS (segmenting pulses into shorter micropulses) has been proposed to mitigate ion back-attraction and promote working gas recovery. This study investigates how micropulse length, delay time between segments, and magnetic field strength influence energy flux, deposition rate, and ionized flux fraction in chopped and standard HiPIMS. These quantities are evaluated by passive thermal probe, biasable QCM and mass spectrometer measurements at the substrate position. Deposition-averaged and pulse-averaged power is kept constant for all conditions to facilitate meaningful comparison. Results indicate that chopping the HiPIMS pulse consistently leads to higher energy flux and total deposition rate compared to standard HiPIMS at the same total pulse length, primarily due to increased ion flux. A weaker unbalanced magnetic field configuration enhances deposition rates and ion transport. In chopped HiPIMS, increasing micropulse length decreased energy flux and deposition rates, whereas increasing the delay time between micropulses substantially improved these parameters. Importantly, standard HiPIMS, which operated at higher frequencies and short pulse lengths, demonstrated superior performance (with higher total energy and particle fluxes) than chopped HiPIMS when compared at similar short pulse durations. This suggests that consistent short pulse durations and sufficient off-times for complete gas refill are paramount for maximizing ion fluxes and deposition rates.

Figures (6)

• Figure 1: (a) Schematic of the experimental setup with a mass spectrometer (MS). (b) Passive thermal probe with an integrated thermocouple used for temperature measurements and a separate for external bias connection (left floating). (c) QCM ion meter with an external magnetic field, which prevents electrons from reaching the QCM crystal. The mass spectrometer head has been raised to accommodate either the thermal probe or the QCM ion meter at the same distance from the target.
• Figure 2: Selected waveforms of the target voltage, the target current, and the power for standard HiPIMS and various chopped HiPIMS pulse configurations (5($S_{\mathrm{1}}$, $S_{\mathrm{O}}$, $D$)). The average power and pulse power were kept the same across all conditions.
• Figure 3: (a) Energy flux measured by the thermal probe at floating potential, (b) neutral Ti atoms, Ti ions, and total deposition rates obtained from the QCM ion meter, and (c) integrated ion energy flux distribution functions over the energy range 0–20 eV measured by MS, for various pulse modes, standard ($130\,\mu\mathrm{s}$) and chopped HiPIMS 5(50, 20, 40)) under different magnetic field configurations. The average power and pulse power were kept the same for all conditions.
• Figure 4: (a) Energy flux measured by the thermal probe at floating potential, (b) neutral Ti atom, Ti ion, and total deposition rates measured by the QCM ion meter, and (c) integrated ion energy flux distribution functions over the energy range 0–20 eV measured by MS, for various chopped HiPIMS pulse configurations. The first and other pulse lengths ($S_{\mathrm{1}}$, $S_{\mathrm{o}}$) were increased by $10\,\mu\mathrm{s}$ in each step. The average power and pulse power were kept the same for all conditions.
• Figure 5: (a) Energy flux measured by the thermal probe at floating potential, (b) neutral Ti atom, Ti ion, and total deposition rates obtained from the QCM ion meter, and (c) integrated ion energy flux distribution functions over the energy range 0–20 eV measured by MS, for various pulse configurations, standard ($130\,\mu\mathrm{s}$) and chopped HiPIMS 5(50, 20, D). The delay time between segments (D) was increased by $10\,\mu\mathrm{s}$ in each step. The average power and pulse power were kept the same for all conditions.