Time-dependent studies of pulsed inductively coupled Ar and Ar/CF4 discharges are presented in this work. By using a time-resolved power diagnosis system, i.e., a Langmuir probe and a Hairpin probe, the temporal evolutions of input power and electron density are measured. In the initial pulse stage, the input power exhibits two peaks, which are related to the properties of the source and the plasma, respectively. In addition, an overshoot of the electron density is observed in the initial pulse stage at high powers (500–800 W) and low pressures (1–10 mTorr), and the overshoot becomes weaker by increasing pressure (10–80 mTorr) or decreasing input power (200–500 W). This can be explained by the dependence of the power transfer efficiency on pressure and input power, as well as the balance between the electron production and loss rates. When the power is turned off, the electron density and the input power exhibit a peak at the initial afterglow period, due to the release of charges from capacitors and inductors in the radio frequency power source. In Ar/CF4 discharges, the plasma responds to the changes in the input power more quickly than in Ar discharges, so it takes a shorter time to reach the ionization equilibrium. This may be caused by more ionization channels, larger ionization cross section, and lower ionization thresholds in Ar/CF4 plasmas.

Pulse inductively coupled plasma has been widely used in the microelectronics industry, but the existence of overshoot phenomenon may affect the uniformity of plasma and generate high-energy ions, which could damage the chip. The overshoot phenomenon at various spatial locations in pulsed inductively coupled Ar and Ar/CF4 discharges is studied in this work. The electron density, effective electron temperature, relative light intensity, and electron energy probability function (EEPF) are measured by using a time-resolved Langmuir probe and an optical probe, as a function of axial and radial locations. At the initial stage of pulse, both electron density and relative light intensity exhibit overshoot phenomenon, i.e., they first increase to a peak value and then decrease to a convergent value. The overshoot phenomenon gradually decays, when the probe moves away from the coils. Meanwhile, a delay appears in the variation of the electron densities, and the effective electron temperature decreases, which may be related to the reduced strength of electric field at a distance, and the consequent fewer high-energy electrons, inducing limited ionization and excitation rate. The overshoot phenomenon gradually disappears and the electron density decreases, when the probe moves away from reactor centre. In Ar/CF4 discharge, the overshoot phenomenon of electron density is weaker than that in the Ar discharge, and the plasma reaches a steady density within a much shorter time, which is probably due to the more ionization channels and lower ionization thresholds in the Ar/CF4 plasma.

The pulsed inductively coupled plasma (ICP) has considerable potential to satisfy multiple stringent scaling requirements for use in the semiconductor industry. However, overshoot of plasma parameters during the rising period of the pulse affects the stability and uniformity of the plasma and can lead to a breakdown of the wafer and over-sputtering of the film. In this study, a step waveform modulation method is used to reduce the overshoot at the initial stage of the pulse. The behavior of the discharge is monitored by measuring (i) the modulated step waveform signal on the function generator, (ii) the input power (by a time-resolved VI-probe), and (iii) the amplitudes of the coil voltage and current (by voltage and current probes, respectively), as well as (iv) the plasma parameters including the electron density, the effective electron temperature, and the electron energy probability distribution function (by a time-resolved Langmuir probe). It was found that the state of the plasma can be controlled by changing the waveform, such as varying the time of the rising edge, varying the initial amplitude, and varying the duration of the low-high amplitude. The results indicated that the overshoot value of the electron density can be reduced by using a low-high step waveform. When the amplitude of the waveform was 500/550 mV and the duration was 200/300  μs, the overshoot value observed was 1/4 of that of the conventional ICP pulse discharge. In addition, increasing the duty cycle of the pulse could also reduce the overshoot value due to the high electron density that occurs during the afterglow period. Moreover, the plasma can reach a steady state more quickly at high pressure by using a step waveform of high amplitude.