Radio Frequency Metrology for Mobile Atmospheric Pressure Plasma Devices

Transcription

1 PIERS ONLINE, VOL. 4, NO. 5, Radio Frequency Metrology for Mobile Atmospheric Pressure Plasma Devices V. J. Law, N. O Connor, and S. Daniels National Center of Plasma Science and Technology, Dublin City University Collins Avenue, Glasnevin, Dublin 9, Dublin, Ireland Abstract Real-time time-domain and frequency-domain measurement of mobile atmospheric pressure plasma is described. The measurements are non-invasive and deployed on the main power-line to the plasma device. 1. INTRODUCTION The atmospheric pressure coaxial dielectric barrier discharge (DBD) has been established for nonthermal plasma processing for more then 2 years [1, 2]. Their potential for direct impact on society in the guise of surface modification of engineering materials and the destruction of microbial pathogens on contaminated surfaces are now becoming apparent. Recently atmospheric pressure glow discharges in the form of: plasma pencil [3] needle [4, 5], jet [6] and torch [7] have been reported to have low electrical power consumption. The common feature of these devices is that they have a volume discharge that acts as a source of ion species and an expanding plume that is driven by the flow of gas passing through the chamber volume. The drive circuitry typically employs a variety of sinusoidal or pulse power technology; some of which have fixed frequencies (13.56 MHz [5, 6]) or variable frequencies (7 khz and 1 MHz [7 9]). Within this body of work and within the water and food purification sectors [1], it is commonly believed that continuous power switching technology has better power conversion efficiencies than sinusoidal excitation. This paper presents the application of Radio Frequency (RF) metrology to atmospheric pressure plasma devices [11, 12]. In particular, to semiconductor power switching circuitry and Flyback transformer that drives an atmospheric discharge. Section 2 describes active RF metrology deployed on the power-line of the Flyback driven DBD. Section 3 describes passive RF metrology on the power-line. Section 4 describes the combination of time-domain and frequency-domain passive RF metrology of a sinusoidal driven plasma. Finally Section 5 provides a conclusion to this work. 2. ACTIVE RF METROLOGY The salient points of the helium DBD have been presented in [12, 13]. Here only the RF measurement is described. The drive circuitry and the Rohde & Schwarz FSL3 scalar network analyzer is shown in Figure 1. Figure 2 shows the equivalent electrical model (EEM) of the measurement. The EEM comprises a 7 element Chebyshev filter front-end probe, a 1.2 m length of 5-Ohm transmission-line and a variable resistor termination (R 3 ) that represents the plasma. Figure 3 shows the measured frequency dependant modulus of the plasma as a function of power: the resolution bandwidth (RBW) is set to 3 khz. Here it can be seen that the modulus is plasma dependent at 165 MHz. A standard linear software package is used to simulate these measurements and after a few parameter iterations the transmission-line responds to the plasma termination at 165 MHz, see Figure 4. At this frequency the EEM yields a resistance termination of 1 21 Ohm with a transmission length of l = 1.3 m and a velocity factor of.62. N.B. a relative offset of 23 db is used to normalize the dynamic gain and account for the insertion of the Chebychev filter. The additional l =.1 m and reduced velocity factor may also be attributed to the additional line length through filter. These simulations show the depth of the node increasing with applied power, thus indicating the helium DBD is becoming more conductive as the power is increased. From a plasma point-of-view this is expected as the ion/electron density increases with applied power. 3. PASIVE RF METROLOGY In this section the same helium DBD is integrated using passive RF metrology to measure the drive oscillator phase noise and frequency pulling as a function of plasma plume-surface interaction [12].

2 PIERS ONLINE, VOL. 4, NO. 5, Hybrid transformer R&S FSL3 High pass filter He DBD Plume Transmission -line Pick-up coil Secondary Primary 2N355 Sweep In 5R 5R Out High-pass filter Transmission-line R3 Ground 1-1 Vdc 2 Amp 24 R 27 R Figure 1: Schematic of SNA and DBD W W 1 W W 17 W W 23.4 W E E+8 1.5E E+8 2.E+8 Figure 3: Reflection modulus of DBD as a function of power. Figure 2: EEM of DBD circuit: 5 Ohm transmission = 1.5 m in length R3 = 21 Ohm R3 = 1 Ohm W 23.4 W E+8 1.2E+8 1.4E+8 1.6E+8 1.8E+8 2.E+8 Figure 4: EEM simulation of DBD. l = 1.3 m, velocity factor =.62, and R3: 1 and 21 Ohm. The measurement is made on the power-line as shown in Figure 1. However, in this case the spectrum analyzer is protected from the high voltage input level by a capacitive clamp front-end probe which has a voltage coupling factor of.1. To exemplify this measurement, Figures 5 and 6 shows the helium DBD driven at 177 KHz with an applied dc power of 11 W under stable and unstable modes of operation due to flow rate conditions. The terms stable and unstable are used to classify visual instabilities within the plasma plume. Here the only operating difference between the stable and unstable mode is the gas flow rate: 2 standard liters per minute (slm) for Figure 5 and 5 slm for Figure 6. These results show that the spectrum analyzer captures the flow rate instability as a 2 khz modulation of the drive carrier frequency. This instability is in the audible range of the human ear and can be clearly heard and captured on a microphone. In the second example the oscillator drive frequency is measured as a function of plasma plume interaction with two different surfaces (copper connected to ground and glass) at a distance of 16 mm, and plume expanding into free-space. The helium flow for these measurements is 2 slm at a drive frequency of khz and applied dc power of 1.5 W. The results for the copper (dotted-line) glass (dashed-line) and air (solid-line) are shown in Figure 7. The result shows frequency pulling of 7 Hz from free space to metal coupling with the plume. For example: khz without

3 PIERS ONLINE, VOL. 4, NO. 5, l/m 5 l/m Gain (dbm) E+5 1.7E+5 1.8E+5 1.9E+5 2.E E+5 1.7E+5 1.8E+5 1.9E+5 2.E+5 Figure 5: Stable operating mode. Figure 6: Unstable operating mode. the work-piece; with the glass slide and; khz with the copper strip. This equates to a frequency pulling ratio of f max / f min of 1.4. This result indicates that the plume-surface interaction is adding a reactive component to the circuit. It is thought that this affect may have many processing applications mm glass Cu Gain (dbm) E E+5 1.7E E+5 1.8E+5 Figure 7: Oscillator frequency pulling due plasma plume-surface interaction. RBW = 1 Hz. 4. PULSE CURRENT WIDTH MODULATION In this section a Dow Corning plasma solutions Labline TM reel-to-reel helium atmospheric pressure plasma system [13] is monitored using passive RF metrology. The time-domain voltage and current waveforms are captured and compared to the frequency-domain current signal. The purpose of these measurements is to identify the effective pulse of the current drawn by the plasma in each half of the driving frequency period. The voltage and current waveforms for a 14 W helium plasma is shown in Figures 8 and 9. The spectrum analyzer frequency-domain envelope measurement was optimized by setting the RBW to be larger than the carrier frequency, but significantly smaller than the effective current pulse width, see Figure 1. A comparison between time-domain and frequency-domain measurements reveals a correlation between the envelope nulls and the current pulse width. For example the nulls have a frequency span (t) of 14 KHz which approximate to the current pulse width ( 7 µs). In effect the current

4 PIERS ONLINE, VOL. 4, NO. 5, Voltage Current Voltage Current Voltage (kv) 4.5 Current (A) Voltage (kv) 4 ~.7s.5 Current (A) Time (s) Figure 8: Voltage and current waveforms Time (s) Figure 9: Single current pulse. 1-1 Period t Period T -7.E+ 2.E+5 4.E+5 6.E+5 8.E+5 1.E+6 Figure 1: Frequency-domain envelope measurement of current waveform using a 3 khz RBW. pulses drawn appears to be pulse width modulating the carrier. If this is the case, the micro-scale events (Figure 6:.1 µs with current amplitudes of.15 A) may have an influence on the effective current and hence the power dissipated in the plasma. 5. CONCLUSION This paper has presented active and passive RF metrology measurement of atmospheric pressure plasmas. These measurement are non-invasive (they are only placed on the power-line). Hence they are not limited by the size of the plasma device: This being an important factor when the plasma device is miniaturized for mobile hand-held applications. The active RF Metrology approach provides an indication of the DBD electrical performance, and when coupled with an EEM the scalar measurement provides information on plasma termination. To gain access to the plasma impedance a vector network measurement needs to be implemented. Passive RF metrology measurements do not require an EEM to interpret the results in that they can be directly correlated to gas flow and plume-surface interactions. The reduction in instrumentation complexity when compared to active RF metrology makes this approach attractive. The frequency-domain measurements provide both temporal (instabilities) and spatial (plume-surface interaction) information. Both of these attributes are necessary for monitoring plasma-surface interactions. The combination of time and frequency-domain measurements allows the current drawn by the plasma to be interrogated. Micro-scale plasma events in the current waveform are likely to modify the effective current pulse width and hence the power developed

5 PIERS ONLINE, VOL. 4, NO. 5, within the plasma. The origin and timing of these micro-scale events are of great interest as they may disclose the underlying plasma physics and chemistry. ACKNOWLEDGMENT This work is supported by Enterprise Ireland under grant number CFTD/7/IT/34. REFERENCES 1. Eliasson, B., M. Hirth, and U. Kogelschatz, Ozone synthesis form oxygen in dielectric barrier discharge, J. Phys. D: Appl. Phys., Vol. 2, No. 11, , Chen, Q., Y. Zhang, E. Han, and Y. Ge, Atmospheric pressure DBD gun and its application in ink printability, PSST, Vol. 14, No. 4, , Laroussi, M., Nonthermal decontamination of biological media by atmospheric pressure plasma: Review, analysis, and prospects, IEEE Trans. Plasma Sci., Vol. 3, No. 4, , Kieft, I. E., E. P. Van den Laan, and E. Stoffels, Electrical and optical characterisation of the plasma needle, New J. Phys., Vol. 6, 149, Stoffels, E., I. E. Kieft, R. E. J. Sladek, L. J. M. Van den Bedem, E. P. Van den Laan, and M. Steinbuch, Plasma needle for in vivo medical treatment: Recent developments and perspectives, PSST, Vol. 15, No. 4, S169 S18, Walsh, J. L., J. J. Shi, and M. G. Kong, Contransting characteristics of pulsed and sinusoidal cold atmospheric plasma jets, Appl. Phys. Letts., Vol. 88, No. 17, , Anghel, S. D. and A. Simon, Measurement of electrical characteristics of atmospheric pressure non-thermal He plasma, Meas. Sci. Technol., Vol. 18, , Panousis, E., F. Cléments, J.-F. Loiseau, N. Spyrou, B. Held, M. Thomachot, and L. Marlin, An electrical comparative study of two atmospheric pressure dielectric barrier discharge reactors, PSST, Vol. 15, No. 4, , Kong, M. G. and X. T. Dang, Electrically efficient production of a diffuse nonthermal atmospheric plasma, IEEE Trans. Plasma Sci., Vol. 31, No. 1, 7 18, Rowan, N. J., S. J. MacGregor, J. G. Anderson, D. Cameron, and O. Farish, Inactivation of mycobacterium paratuberculosis by pulsed electric fields, Appl. Environ. Microbiol, Vol. 67, No. 6, , Law, V. J., F. J. Lawler, and S. Daniels, Non-invasive VHF Injection signal monitoring in atmospheric pressure plasma and DC Magnetron, Vacuum, Vol. 82, No. 5, , Law, V. J., Process induced oscillator frequency pulling and phase noise within plasma systems, Vacuum, 82, doi:1.116/j, vacuum Twomey, B., D. P. Dowling, G. Byrne, L. O Neill, and L. O Hare, Properties of siloxane coatings deposited in a reel-to-reel atmospheric pressure plasma system, Plasma Process and Polymers, Vol. 4, , 27.