Abstract
An alternating current (AC) gliding arc can be conveniently operated at atmospheric pressure and efficiently elongated into the ambient air by an air flow and thus is useful for surface modification. A high speed camera was used to capture dynamics of the AC gliding arc in the presence of polymer surfaces. A gap was observed between the polymer surface and the luminous region of the plasma column, indicating the existence of a gas boundary layer. The thickness of the gas boundary layer is smaller at higher gas flow-rates or with ultrasonic irradiation to the AC gliding arc and the polymer surface. Water contact angle measurements indicate that the treatment uniformity improves significantly when the AC gliding arc is tilted to the polymer surface. Thickness reduction of the gas boundary layer, explaining the improvement of surface treatment, by the ultrasonic irradiation was directly observed for the first time.
Keywords
Introduction
Non-thermal plasma processing at atmospheric pressure is widely used for surface modification.1,2 Its applications include surface cleaning, 3 decontamination and sterilization,4,5 deposition of functional coatings 6 and improvement of adhesion, wetting and paintability.7–11
One of the challenges in non-thermal plasma processing at atmospheric pressure is to achieve high reactivity and high productivity simultaneously. 12 Here, high reactivity is ensured by a high electron temperature in a non-equilibrium plasma so as to change chemical bonding of molecules in the plasma or the plasma treated surfaces. Meanwhile, high productivity can be demonstrated using high energy densities. However, most plasmas can hardly sustain a non-equilibrium state at high energy densities. Two possible approaches are proposed to overcome this issue: (1) development of a hybrid plasma with a high energy density in a non-equilibrium state and (2) plasma processing assisted by an external energy input.
A gliding arc is one of the hybrid plasmas, 13 generated between diverging electrodes as a low impedance thermal arc discharge, extended by a gas flow and quenched to non-thermal condition. 14 It can be operated in atmospheric pressure air and thus advantageously used for large-scale processing.15–20 An alternating current (AC) gliding arc has a long lifetime extending over hundreds of AC periods without extinction. The plasma column can be elongated to approximately 20–30 cm.16–20 The AC gliding arc is useful for adhesion improvement of glass fibre reinforced polyesters (GFRPs),16,17 efficiently oxidising GFRP surfaces when the distance between the edge of the electrodes and the GFRP surface is up to 6 cm in open air. It is also demonstrated that optical techniques are promising for non-intrusive diagnostics of the AC gliding arc.18,19 In particular, the dynamics of the gliding arc is observed by using a high speed camera. However, high speed observations of the AC gliding arc in the presence of a surface have not been intensively studied.
The efficiency of plasma processing can be improved by simultaneous high power ultrasonic irradiation to the plasma and material surfaces.21–27 It is reported that high power ultrasonic irradiation during the gliding arc surface treatment enhances oxidation of GFRP surfaces. It has been proposed that a gas boundary layer sticking at a material surface reduces its thickness when the surface is irradiated with acoustic waves. Thus, the turbulent fluxes of the reactive species towards the surface increase and the surface modification is enhanced. However, this reduction of the boundary layer thickness in plasma processing was never visually demonstrated.
In the present work, the AC gliding arc is generated for surface treatment of GFRP plates. The performance of the AC gliding arc directing is captured by a high speed camera and OH planar laser induced fluorescence (OH-PLIF), while the surface modification effect is monitored by static water contact angle measurements and X-ray photoelectron spectroscopy (XPS). The effects of gas flow-rates, a tilted gliding arc and ultrasonic irradiation to the AC gliding arc are investigated.
Experimental methods
The AC gliding arc was generated between two diverging tubular electrodes. The outer diameter of the tubular electrodes is 3 mm. Cooling water was fed through the electrodes during operation.16–20 An air flow was fed between the electrodes with flow-rates of 14·0, 17·5, 21·0, 31·5 or 42·0 standard liters per minutes (SLM). The AC gliding arc was driven by an AC power supply at a frequency of 31·25 kHz (Generator 6030; SOFTAL Electronic GmbH). Two millimeter thick GFRP plates were used as specimens (G-Etronax PM; Elektro-Isola). They were cleaned and degreased with acetone and methanol before the plasma treatment.
In order to treat a GFRP plate surface without excess thermal damage from the AC gliding arc, a poly(methyl-methacrylate) holder was moved forward and backward at a speed of 180 mm s−1 on which the GFRP plate was fixed. The angle between the gas flow direction and the specimen surface is approximately 30° (also referred to as “tilted”) or 90° (also referred to as “vertical”). A line formed by the specimen surface and a plane containing the electrodes is perpendicular to the moving direction of the specimen, unless mentioned otherwise. Experimental arrangements of the specimen and the gliding arc are schematically illustrated in Fig. 1.
Experimental arrangements of specimen and AC gliding arc: angle between the gas flow direction and specimen surface is either 90° (vertical: a, b) or 30° (tilted: c, d); arrangement between the moving direction of specimen and the line formed by the specimen surface and the plane containing electrodes is either perpendicular (a, c) or parallel (b, d)
For the investigation of the effect of the ultrasonic irradiation, the angle between the gas flow direction and the specimen surface was fixed at approximately 30° and the acoustic waves were introduced vertically to the specimen surfaces through a cylindrical waveguide using a high power gas jet ultrasonic generator (SonoSteam; FORCE Technology). The acoustic frequency range is between 20 and 40 kHz, and a sound pressure level is approximately 150 dB.21–27 A high-speed camera (Phantom v7·1) was used to capture the behaviour of the AC gliding arc during surface treatment. The high speed camera was operated at 20 kHz with exposure time of 30 μs and resolution of 256×256 pixels in the measurements.
PLIF is a powerful method for specific species detection and is widely used in the investigation of combustion 28 and plasma. 29 In order to visualize the spatial distributions of the reactive species generated by the AC gliding arc, OH-PLIF was performed in different directions. A Nd:YAG (Brilliant b) pumped dye laser (Continuum ND60) was tuned to 283·268 nm to achieve excitation of OH X2 Π–A2 Σ+ (0,1). The resulting fluorescence at around 308 nm was collected by an ICCD camera (Princeton PI-MAX II) with an exposure time of 30 ns. Filters (UG5+WG295) were added in front of the camera to reduce the laser scattering.
The static contact angles for deionized water on the GFRP surfaces were measured in air at room temperature using a contact angle measurement system (CAM100; Crelab Instruments AB).
XPS data were collected using a microfocused, monochromatic Al Kα X-ray source with a lateral resolution of 30 μm (K-alpha; ThermoFischer Scientific, UK) to study the changes of the elemental composition at the GFRP surfaces. Atomic concentrations of each element were calculated by determining the relevant integral peak intensities subtracting a linear background, and the O/C ratio was obtained.
Results and discussion
Effects of flow-rate and distance to surface
The angle between the gas flow direction and the GFRP surface is fixed at approximately 90° and the distance from the head of the electrodes to the specimen surface is 15 mm during the treatment. Averages of 10 000 images taken by the high speed camera and the temporal photoemission intensity from the projected squared area (0·5 mm high and 10 mm wide) near the specimen surface are shown in Fig. 2. The photos of the AC gliding arc in Fig. 2 indicate that a gap exists between the GFRP surface and a luminous region where the AC gliding arc is likely present. Here, the gap is different from a plasma sheath in that the AC gliding arc can be sustained regardless of the existence of the GFRP surface. Instead, the observed gap represents a gas boundary layer which is always formed close to the surface due to viscosity. The discharge tends to approach the specimen surface at higher flow-rates. The time distribution of the photoemission in the region just above the surface in Fig. 2 shows that the overall photoemission intensity tends to increase as the gas flow-rate increases. The tendency of the photoemission of the AC gliding arc at different flow-rates is similar to the observation without a surface. 17
Average of 10 000 images (left panels) and temporal photoemission intensity near the specimen surface (right panels) of the AC gliding arc at different flow-rates: a 14·0; b 17·5; c 21·0; d 31·5; e 42·0 SLM
The increasing overall photoemission at higher flowrates results from more frequent conversions from glow type gliding discharge to spark type. 19 It is therefore indicated that the gas boundary layer can be detected by the high speed camera, and that the thickness of the gas boundary layer decreases at higher flow-rates. The result would suggest that at higher gas flow-rates, surface treatment effects would be improved due to a thinner gas boundary layer between the plasma and the surface. However, it is found in Fig. 2 that the photoemission intensity fluctuates more at higher flow-rates, affecting the treatment uniformity. Moreover, it is reported that as the gas flow-rate increases, the concentration of hydroxyl radical decreases, which is thought to be among the most important oxidative agents for plasma treatment, 19 and that the wettability of the treated GFRP surfaces decreases due to the excess quenching of the plasma. 16 In fact, when the flow-rate is higher than an optimal flow-rate, the photoemission intensity frequently goes down to zero, corresponding to a frequent extinction of the AC gliding arc by quenching. It is therefore expected that there is an optimum gas flow-rate for the gliding arc surface modification. Utilizing the same set-up and conditions, the optimum gas flow-rate of 17·5 SLM is reported in terms of the wettability as shown in Table 1, which agrees well with the previous work since the gas boundary layer at 17·5 SLM is modest and the intensity fluctuation is relatively small.
Water contact angles of GFRP plates after gliding arc treatment at different flow-rates: result is based on Ref. 17 (© IOP Publishing. Reproduced by permission of IOP Publishing. All rights reserved.)
Figure 3 shows averaged images of the AC gliding arc when the distances between the specimen surface and the edges of the electrodes are 15 and 65 mm. The photoemission intensity drops down when the distance increases. Similar results were reported in that at higher flow-rates, the wettability and the degree of oxidation16,17 at the GFRP surface decrease and that the photoemission intensity decreases. The oxygen/carbon (O/C) ratio measured by XPS and the water contact angle of the GFRP plates after the gliding arc treatment are summarized in Table 2. The result indicates that the GFRP surface can be efficiently oxidized when the distance between the specimen surface and the edges of the electrodes is up to approximately 6 cm, and that the oxidation effect significantly decreases when it is larger than approximately 6 cm. More detailed surface characterization and dedicated adhesion tests after the treatment are reported by Kusano et al.15–17
Average of 10 000 images of AC gliding arc: distances of edge of the electrodes to the specimen surface are a 15 mm and b 65 mm
O/C ratios and water contact angles of GFRP plates after gliding arc treatment at different distances: result is based on Ref. 17 (© IOP Publishing. Reproduced by permission of IOP Publishing. All rights reserved.)
Tilted gliding arc
When a gliding arc is used for surface treatment, the gas flow direction is usually perpendicular to the surfaces due to a short extension of the general gliding arc reported in a literature.30,31 That is, the gliding arc is usually directed perpendicular to the surface to be treated. However, even when the angle between the gas flow direction and the specimen surface is not perpendicular, the AC gliding arc in the present work can extend long enough (up to 20–30 cm at 17·5 SLM) to approach the surface. The static water contact angles were measured at four different conditions at the fixed distance of 15 mm between the edges of the electrodes and the specimen surface. The gas flow-rate was 17·5 SLM. The power to the discharge was approximately 800 W. Each specimen surface was exposed to the discharge twice. The angle between the gas flow direction and the specimen surface was either 30° (tilted as shown in Fig. 1c and d) or 90° (vertical as shown in Fig. 1a and b). The arrangement between the moving direction of the specimen and the line formed by the specimen surface and the plane containing the two electrodes was either perpendicular or parallel. The results are shown in Fig. 4.
Water contact angles of GFRP surfaces after AC gliding arc treatment: see Fig. 1 for conditions
When the angle between the gas flow direction and the specimen surface is 90° (vertical), measured contact angles were scattered, indicating a poor treatment uniformity by the gliding arc. On the other hand, when the tilting angle was 30° (tilted), approximately 4 cm width was relatively uniformly and well treated, and thus, the tilting appears useful for improving the treatment uniformity. One possible explanation of the improvement is less fluctuation of the discharge column due to less streamline curvature and less shear in the tilted case. Furthermore, even when the arrangement between the moving direction of the specimen and the line formed by the specimen surface and the plane containing the two electrodes is parallel (Fig. 1d), the surface was uniformly treated within a certain region as shown in Fig. 4d. It is therefore indicated that the discharge column can dynamically fluctuate in a direction perpendicular to the plane containing the electrodes.
The OH-PLIF images of the AC gliding arc without the GFRP plates were collected to investigate the distribution of ground state OH radicals generated by the plasma. As shown in Fig. 5a, the asymmetric OH-PLIF image parallel to the plane containing the electrodes indicates that the plasma is partly out of plane of the laser sheet and is three-dimensionally distributed. It is confirmed in Fig. 5b that the plasma is also distributed perpendicular to the plane containing the electrodes. The observation in Fig. 5 shows good agreement with the result that the GFRP surface was uniformly treated even when the arrangement between the specimen moving direction and the line formed by the specimen surface and the plane containing the electrodes is parallel as shown in Fig. 4d. The relevant dynamics of the gliding arc discharge column is studied elsewhere. 19
Average of 500 OH-PLIF images of the AC gliding arc taken by the ICCD camera a facing and b perpendicular to plane containing electrodes: the laser sheet was located at plane parallel to OH image
Effect of ultrasonic irradiation
The advantage of tilting the AC gliding arc is not only the uniform treatment, but also easy introduction of ultrasonic irradiation. It is reported that the treatment effects of the AC gliding arc are enhanced by the ultrasonic irradiation. 25 In the present work, high speed images were taken with and without ultrasonic irradiation. The gas flow-rate was 17·5 SLM, and the power to the gliding arc was approximately 800 W. A distance between the edges of the electrodes and the specimen surfaces was 15 mm. Without ultrasonic irradiation, a gap of a few millimeters was observed without strong photoemission between the surface and the photo emitting part as shown in Fig. 6a. However, with ultrasonic irradiation, the gap significantly reduced at the same flow-rate. This is the first direct observation of the thickness reduction of the gas layer attaching at the surface in plasma surface processing due to ultrasonic irradiation. The result corresponds to the improved plasma treatment effects by the ultrasonic irradiation21–27 in which detailed surface characterizations are reported. More specifically, the surface density of the carboxyl group seems to increase with ultrasonic irradiation. 25 It is also noted that the ultrasonic irradiation induced observed spatial fluctuation and more frequent shortcutting events than the AC gliding arc without ultrasonic irradiation, and that the subsequent averaged image in Fig. 6b looks as if the discharge column would be discontinued.
Average of 10 000 images of the tilted gliding arc at gas flow-rate of 17·5 SLM b with and a without ultrasonic irradiation
Conclusion
The performance of the AC gliding arc surface treatment was investigated using a high speed camera. The photographs revealed that the thickness of the gas boundary layer tended to decrease as the flow-rate increased so that the plasma column was observed closer to the surface on average. On the other hand, the photoemission intensity of the plasma fluctuated more and affected the treatment uniformity. The results indicate the existence of an optimum gas flow-rate for the surface treatment. The long extension of the discharge enables demonstration of plasma surface treatment with the tilted gliding arc. It is shown that the tilting improved uniformity of the treatment. For example, when the tilting angle was 30°, approximately 4 cm width of the specimen surface was relatively uniformly treated. Ultrasonic irradiation to the gliding arc also reduced the thickness of the gas boundary layer, indicating improvement of surface treatment effect. The effect of the thickness reduction by the ultrasonic irradiation, which has previously been predicted, was directly observed for the first time. It is concluded that direct observation of the AC gliding arc helps understanding of dynamics of plasma surface processing.
Acknowledgements
The work had a partial financial support from the Swedish Energy Agency, Swedish Research Council (VR) and Knut & Alice Wallenberg Foundation. J.J. Zhu would like to thank China Scholarship Council for financial support.