A comparative study between pulsed-DC and RF plasma pre-treatment of polymer web
Plasma surface treatment of web material is an optional way to improve the adhesion, which is important for barrier properties, electrical performance, or other desired properties. This paper presents an experimental comparison study of RF and pulsed DC glow discharge plasma pre-treatment of orientated polypropylene (OPP) web. The aim is to improve the web surface energy, which should enhance primarily the adhesion and consequently improve the electrical property of metalized web as an important point in the fabrication of capacitors. For commercialization purposes, the influence of N2 and N2/O2 plasma gas on the web surface modification was investigated. For the surface analysis, contact angles were used to evaluate the surface energy (SE) of OPP web induced by both discharge systems. The pre-treatment using pulsed DC with 100 sccm N2 in combination of 50sccm O2 showed values up to 36 kJ/cm2 surface energy induced on the web as compared to 29 kJ/cm2 for untreated films. These values were confirmed by carrying out the dyne test according to ASTM-D2578. This increase in surface energy effect proved to be advantageous for the production of capacitors.
The purposes of plasma pre-treatment of webs are namely for cleaning, sealing, etching, and functionalization.[2-4] For the past 20 years, pre-treatment in the vacuum web coating technology has become an indispensable tool for diverse applications, e.g. to enhance the adhesion of metalized films on polymeric packaging and capacitor web. For the latter applications the electrical characteristic is important, which can be influenced by choice of treatment. This treatment has commonly been carried out by atmospheric corona discharges, flame or by glow discharges under low pressure (mbar range).
These surface treatments have a significant influence on the film growth. For capacitor design, the adhesion of vacuum metalized films on polymer surface should be maximized in order to improve the electrical properties desired. Improving the adhesion of the subsequent vacuum metalized film yields to a better film growth consequently enhanced mechanical stability. During capacitor fabrication the metalized films undergo stress during the winding process, which might lead to structural defects. These defects are detrimental to a capacitor’s electrical performance, hence improving the adhesion by optimizing the surface energy would yield mechanically stable films. The optimal web surface treatment can only be determined for individual treatment stations, and their results are not transferable. Thus, the operating parameters of each experiment must be examined to determine the surface energy.
In this present work, two separate plasma gas mixtures N2 and N2/O2 were used in an RF and a pulsed DC magnetron discharge to enhance the web surface energy. The surface energy was calculated from the contact angles values determined from different test liquids used. The dependencies of surface energy property with respect to the discharge input power and gas flow rate are presented and discussed in terms of mechanisms occurring in the plasma and the counteraction between the plasma and web surface.
Furthermore, the electrical properties of capacitors fabricated from films pre-treated using (N2 and N2/O2) RF and pulsed DC discharge have been compared.
Web treatment with magnetron plasma:
Plasma discharges have been implemented in various sectors in the thin film industry. In addition to applying a DC or RF potential difference (or electric field) to a glow discharge, a magnetic field can also be applied. The well established type of discharge is the crossed magnetic and electric field, known as magnetron discharge. Different configurations can be distinguished, cylindrical, circular and planar magnetrons.[6, 7]
Magnetically enhanced pretreatment processes have proven to enhance adhesion and barrier performance of web, depending upon an optimum dose. However, the effect of enhancing the adhesion property of the web for capacitor films by applying different voltages has so far not been reported.
When a constant potential difference is applied between the cathode and anode, a continuous current will flow through the discharge; giving rise to a direct current (DC) glow discharge.[8, 9] Firstly, to sustain a direct current glow discharges the electrodes have to be conducting. When one or both of the electrodes are non conductive, the electrodes become gradually covered with insulating material during plasma processes. The electrodes will be charged up due to the accumulation of positive and negative charges, and the glow discharge will neutralize each other. This problem has been tackled by applying a time-varying potential difference[10, 11], i.e;
- by applying an alternating voltage (RF, 13.56 MHz) between the two electrodes, so that each electrode will act as the cathode and anode, and the charge accumulated during one half-cycle will be partially neutralized by the opposite charge accumulated during the next half-cycle, or
- the DC voltage can be applied in form of discrete pulses, typically with lengths in order of milli- to microseconds. This can be regarded as a short DC glow discharge, followed by a generally longer afterglow, in which the discharge burns out before the next pulse starts.
With respect to both options, pulsed DC discharges operate at higher peak voltages and peak currents than in the RF glow discharge, thus higher instantaneous ionization and excitation can be expected using DC discharges. This is basic plasma phenomena, whereby excitation and ionization are highly non-linearly dependent on field strength, thus reduced charge induced damage on substrate can be achieved upon using a very short duty cycle, i.e. the ratio of ‘pulse-on’-period compared to ‘pulse-off’ period is very small. More concretely, the average electrical power is low, so that the chamber will not excessively be bombarded.
Applying both options to pre-treat web, the efficiency of both methods should be investigated with respect to the surface modification effect induced using different plasma gases.
Experiments were carried out using a planar magnetron source, which has been designed to treat the moving web, housed in a CAP-M 650 (Leybold Optics GmbH) as shown in Fig. 1. The plasma discharge enhanced by a closed loop tunnel shaped magnetic field is generated in the 10-3 mbar pressure range. The two planar magnetron cathodes are cooled by water, could treat web widths up to 650 mm, and attain web speed up to 18 m/s. They are operated in a typical input power dosage range up to 90 W·s/m2. Operating gases were admitted through leak valves for N2 at flow rates 100 sccm, 200 sccm and 300 sccm, and where stated, O2 gas flow rate was always held constant at 50 sccm for all experiments. The maximum pressure attained in the winding chamber was 10-2 mbar range.
The RF (13.56 MHz) power supply, Seren IPS AT20, was connected to the magnetron cathode via a matching unit. The DC discharges were maintained between the outer planar electrode earthed and an inner electrode powered by pulsed DC power supply (Advanced Energy DC pinnacle plus, 40 kHz @ 1µs). The strength of the magnetic field was constant for all experiments at 113 A/cm. The distance between the magnet assembly and the target was fixed. Figure 1 shows a schematic representation of the magnetron pre-treatment station used for the web processing in this comparative study. The main difference in this comparative study is the power supply used; that is comparing the plasma generated from the RF power supply and matching unit versus pulsed DC power supply.
Contact angles of non-metallized 4.0 µm orientated polypropylene (OPP, Treophan) were measured by the static and dynamic sessile drop technique using a DSA 10 goniometer (Kruss GmbH, Germany). MilliQ-H2O, n-Hexadecane and 1,5 Pentandiol were used as test liquids with a drop volume of 10 µl. All values are the average of five measurements per sample. The Owens-Wendt approach was used to calculate the surface free energy (kJ/cm2) from the contact angle values measured. The dyne test (ASTM-D2578) was used to confirm the calculated values.
4.1Effect of RF glow discharge on web using only N2 process gas:
Figure 2 indicates the effect of using only N2 process gas in the RF glow discharge on the web surface energy. The surface energy generally increases with RF input power, but decreases with gas flow rate. The 0 W indicates the no treatment was carried out on web, i.e. virgin web.
4.2Effect of RF glow discharge on web using N2 and 50 sccm O2 process gas:
Similarly to Figure 2, oxygen gas was used additionally for the web pre-treatment. The surface energy of the modified web by RF glow discharge using N2 and O2 process gas are shown in Figure 3. It can be seen from Figure 3 that the surface energy of the web increases with input power and also decreases with increasing gas flow rate of N2 gas. The O2 gas flow rate was constant for all the experiments. The maximum surface energy induced on the web was 34 kJ/cm2 attained at 1000 W and 100 sccm, input power and N2 gas flow rate respectively. The effect of adding oxygen gas for the pre-treatment showed an increase in surface energy of the web over 2 kJ/cm2 as compared to N2 gas plasma treatment.
4.3Effect of pulsed DC glow discharge on web using N2 process gas:
In Figure 4, only N2 process gas was used for the DC glow discharge treatment. Figure 4 portrays the surface energy is relatively constant upon increasing input power. However, increasing the N2 gas flow rate increases the surface energy. The highest surface energy attained was 33 kJ/cm2 at 300 sccm, 1000 W, whereas 29 kJ/cm2 was measured for untreated web.
4.4Effect of pulsed DC glow discharge on web using N2 and O2 process gas:
Similar to conditions used in Figure 4, the additional oxygen gas flow rate of 50 sccm was kept constant. The surface energy of the web increases upon increasing d.c input power, and gas flow rate, as shown in Figure 5. The highest surface energy induced on web was obtained using these parameters, i.e. 1000 W, 300 sccm N2 and 50 sccm O2 as compared in the Figure 6.
Discussions: RF and DC glow discharge on web:
The effect of the pre-treatment using N2, and N2 + O2 gases for both RF and pulsed DC input power showed different surface energy values. This might be related to the surface modifications induced by the oxygen and/or nitrogen plasma on the web. Depending upon the input power level and gas flow rate (N2, and/or N2 + O2) required to initiate these modifications, diverse surface functional groups are created which influence the surface energy as seen in Figure 2-5. Thus, increasing the surface energy would improve the wettability and bondability (interfacial strength) properties. This effect has been reported by Yasuda and co-workers, whereby oxygen functionalities were incorporated in nitrogen-plasma-treated polymer surfaces at specific input power and treatment time.
Firstly, in both cases presented here, RF and pulsed DC power supply with N2 discharge modified the web to enhance hydrophilic surfaces, i.e. higher surface energy, upon increasing input power. This can be explained by two processes occurring at a defined treatment time (constant web speed); simultaneous etching of the polymer surface due to atomic oxygen reaction with the carbon atoms, and the surface formation of oxygen functional groups through the reactions between the active species from the plasma and the surface atoms is responsible for this observation. The competitive nature of both processes reveals the surface modification at the beginning, and then surface etching at a later stage of the treatment might occur.
Comparing the SE of the untreated web and treated web at 400 W input power, an increase of 4 kJ/cm2 was calculated. This increase in SE might be due to the first surface reaction on the web; surface modification. Higher than 400 W input power, surface etching overshadows the surface modification in this case. This indicates less polar groups are introduced on the surface. The gas combination (N2 + 50 sccm O2) initiated higher surface energy for plasma treated films as compared using N2 gas flow only. The addition of oxygen radicals further indicates the modification and formation of groups such as –OH, C=O, –NO, –NH, which are responsible for the decrease in wettability, thus increasing the SE.
A practical example, has been confirmed by a capacitor producer, whereby the electrical properties were compared for capacitors fabricated from untreated web, pre-treated web with RF and pulsed DC discharge. The best results were obtained from (N2/O2) pulsed DC discharge. This further indicates that SE enhances the adhesion, as well as the mechanical strength of the metalized film on the web.
Secondly, comparing both treatment stations, the web treated by RF discharge requires less gas flow rate than DC discharge to increase the web surface energy. A simplified explanation for these opposite trends is due to the fact that;
- electrons in the RF discharge can oscillate in the plasma between the two electrodes,
- increasing the gas flow rate hinders the electron oscillation and as well the ionization, the latter occurs at different time scales.
As a result, increasing the gas flow rate hinders the oscillation of the electrons and favours the ion radical recombination in the plasma. As a result, operating the RF pre-treatment at a low flow rate of 100 sccm N2 induces higher surface energy.
Finally, the SE has been increased using both treatment units as shown in this work, Figure 6. However, the pulsed DC discharge showed higher values as compared to the RF discharge. The average electrical power is rather low (pulsed input power), so that the sample or web will not excessively be heated, neither create charge-induced damage nor etch profile distortion on the web. Another advantage of pulsed DC technology glow discharges compared to RF technology is the simpler method of up-scaling due to reduced impedance matching network and electromagnetic interference problems, and the lower price of power supplies for larger reactors.
The results of plasma pre-treated web using an RF and a DC magnetron discharge using N2, and N2/O2 have been described. Based on the surface modifications induced by both discharges, the contact angles were measured, which was used to calculate the SE of the modified surfaces.
The SE increased upon increasing input power for both discharge systems. N2 and O2 plasma gas induced SE values up to 36 kJ/cm2 as compared to 33 kJ/cm2 using N2 gas only. The pre-treatment using the RF magnetron discharge in order to enhance SE is favored at lower gas flow rates 100 sccm. On the contrary, the pulsed DC discharge initiated the highest SE attained, 36 kJ/cm2, at a flow rate of 300 sccm (N2/O2).
The surface treatment with pulsed DC discharge proved several advantages over the RF discharge web treatment, namely;
- higher SE were attained by treating the surface with N2/O2 plasma, and this led to an improved adhesion of the subsequently vacuum deposited film.
- This adhesion improvement was further confirmed by the final product test. The capacitors fabricated from the DC pulsed discharge surface treated web showed an increase capacitance with respect to RF discharge treatment.
- Pulsed DC web treatment stations are now built in all the Leybold Optics CAP-M machines.
 V. Geitner, Electronicon Kondensatoren GmbH (2008).
 H. Yasuda, Plasma Polymerization, Academic Press, 1985.
 C. A. Bishop, Vacuum Deposition onto Webs, Films, and Foils, William Andrew Publishing, Norwich, NY., 2006.
 A. N. Chifen, A. T. Jenkins, W. Knoll, R. Förch, Plasma Processes & Polymers 4 (2007).
 G. P. Lopez, B. D. Ratner, Plasma deposition treatment and etching of polymers, Academic Press Inc., 1990.
 B. Chapman, Glow Discharge Processes, New York, 1988.
 M. Sugawara, Surface and Coatings Technology 116 (1999) 543-546
 H. Bäcker, J. W. Bradley, P. J. Kelly, R. D. Arnell, Journal of Physics D: Appl. Phys. 34 (2001) 2709-2714.
 B. Fritsche, T. Chevolleau, J. Kourtev, A. Kolitsch, W. Möller, Vacuum 69 (2003) 139-145.
 A. Bogaerts, E. Neyts, R. Gijbels, J. v. d. Mullen, Spectrochemica Acta Part B 57 (2002) 609-658.
 J. W. Bradley, S. K. Karkari, A. Vetushka, Plasma Sources Sci. Technol. 13 (2004) 189-198.
 D. e. a. Owens, Journal of Applied Polymer Science 13 (1969) 1741- 1747.
 F. M. Petrat, D. Wolany, B. C. Schwede, L. Wiedmann, B. A., Surface Interfacial Analysis 21 (1994) 274-282.