Investigating the Plasma Surface Modification of Polystyrene at Low Ion Power Densities Marshal Dhayal, ² Kristina L. Parry, ‡ Robert D. Short, ‡ and James W. Bradley* Department of Physics, UMIST, SackVille Street, Manchester, M60 1QD United Kingdom, and Department of Engineering Materials, UniVersity of Sheffield, Mappin Street, Sheffield S1 3JD, United Kingdom ReceiVed: May 27, 2004 A new grid-separated, two-chamber low-pressure plasma reactor [Dhayal, M.; Forder, D.; R. Short, D.; Bradley, J. W. Vacuum 2003, 70, 67] 1 has been developed to study the surface modification of polymeric materials under low ion power density bombardment. Through external biasing of the grids, the ion flux Γ and energy E of the bombarding plasma ions at the substrate can be controlled independently (within certain experimental limits). Importantly, this can be achieved without changing the VUV component of the plasma. In the present arrangement, ion power densities (qΓE) in the range 0.7-15.5 W m -2 can be achieved. This allows a low power density regime to be explored, not possible in typical glass barrel RF reactors often used for surface modification of polymeric materials, in which ion power densities are typically 10-160 W m -2 . The effect of ion energy and flux on the surface modification of polystyrene (PS) was quantified by the ratio of atomic oxygen to carbon in the post-treated samples exposed to air (measured by X-ray photoelectron spectroscopy, XPS). It was found that, over the whole available range in ion fluxes, (1.7-5 × 10 18 m -2 s -1 ) and energies (1.5-23 eV), the O/C ratios did not change with O/C about 0.1 to 0.12 for plasma exposure times of 90 s. With elimination of the ion species from the processing region of the discharge through grid biasing, the O/C ratios were also found not to change, indication that below a certain threshold of ion power (at least 15.5 W m -2 ) ions play little role in the surface modification. By replacing the grids with a LiF window (cutoff wavelength λ c 104 nm) having a transparency in the VUV region (<200 nm) at least as good as the geometric transmission area of the two grid system we observe again little change in O/C, indicating that the dominant species in production of free radical sites is the VUV, with any excited neutral species and radicals in the plasma playing a minor role. 1. Introduction The low-pressure (0.1-10 Pa) RF plasma treatment of polymeric materials has become an attractive method to modify the surface properties (without affecting bulk material), provid- ing, for instance, enhanced wettability, printability, and adhesion etc. 2 Due to the mobile nature of plasma species, the ions, electrons, neutrals, and free radicals (and photons), plasma treatment can be used on objects of complicated geometry. Despite the burgeoning literature on plasma treatment, 2 there is little agreement on the relative importance of the different mechanisms involved in the plasma-polymer surface interaction. This confusion arises because reactors used are not designed with the systematic investigation of plasma parameters in mind. Widely employed over the past 25 years, in academia and to a lesser extent in industry, has been the tubular glass reactor (see Clark and Dilks 3 ). This plasma reactor design, despite its continuing popularity, has been poorly characterized, and for research purposes, the competing physical phenomena (such as ion and VUV bom- bardment, contributing to surface bond breaking) cannot easily be differentiated from each other in terms of their contribution to the process. The role of ions in plasma treatment has been “hotly” debated, without much serious experimental evidence. Lack of data stems from a poor knowledge of ion fluxes/ energies/identities (in the plasma), although some attempts have been made to measure the energy flux of the ion and VUV species bombarding isolated polymer substrates, for instance in capacitively coupled RF argon plasma. 4 The presence and effect of the VUV component in low- pressure reactors used for surface modification have been studied extensively. For instance, Clark and Dilks 5 showed that in an argon gas RF coupled plasma the most intense VUV lines were at 104.8 and 106.6 nm corresponding to Ar (I) transitions of the neutral atoms (3p 5 4s f 3p 6 ). In the region studied (65- 110 nm) the intensity of these lines was 2 orders of magnitude greater than the total output of the remainder of the spectrum. Furthermore these intensities increased 3-fold as a function of input power (2 f 20W) and maximized at a gas pressure of 0.65 Pa. Egitto and Mateinzo 6 used a series of UV transmitting crystals and XPS valence band analysis to show that photo- ionization took place when photons with energies greater than the first ionization potential of the polymer were absorbed by the sample. Contact angle measurements were used to study surface changes and they found that modification of polyeth- ylene (PE) occurred with photon energies greater than 8 eV and polytetrafluorethylene (PTFE) with photon energies greater than 10 eV. Shard and Badyal 7 compared VUV (λ < 200 nm) radiation from a nitrogen plasma transmitted through a LiF window with an oxygen plasma and found similar modification produced by the two in samples of PS, PE and polyetherether- ketone (PEEK). The VUV treatment of PE has been studied by Holla ¨nder et al. in a number of papers. They observed increases in the PE surface O/C ratio after several minutes of exposure * To whom correspondence should be addressed. E-mail: j.w.bradley@ umist.ac.uk. ² UMIST. ‡ University of Sheffield. 14000 J. Phys. Chem. B 2004, 108, 14000-14004 10.1021/jp0477046 CCC: $27.50 © 2004 American Chemical Society Published on Web 08/20/2004