Influence of the Plasma Sheath on Plasma ... - ACS Publications

Jun 1, 2009 - Laboratory of Biophysics and Surface Analysis, School of Pharmacy, The UniVersity of Nottingham, UniVersity. Park, Nottingham, NG7 2RD, ...
0 downloads 0 Views 299KB Size
J. Phys. Chem. B 2009, 113, 8487–8494

8487

Influence of the Plasma Sheath on Plasma Polymer Deposition in Advance of a Mask and down Pores Mischa Zelzer,† David Scurr,† Badr Abdullah,‡ Andrew J. Urquhart,†,⊥ Nikolaj Gadegaard,§ James W. Bradley,‡ and Morgan R. Alexander*,† Laboratory of Biophysics and Surface Analysis, School of Pharmacy, The UniVersity of Nottingham, UniVersity Park, Nottingham, NG7 2RD, U.K., Department of Electrical Engineering and Electronics, Brownlow Hill, UniVersity of LiVerpool, LiVerpool L69 3GJ, U.K., and Department of Electronic & Electrical Engineering, Rankine Building, UniVersity of Glasgow, Glasgow G12 8LT, U.K. ReceiVed: March 10, 2009; ReVised Manuscript ReceiVed: April 24, 2009

Plasma species that form plasma polymer deposits readily penetrate through small openings and are therefore well suited to coat the interior of porous objects. Here, we show how the size of the cross section of square channels influences the penetration of active species from a hexane plasma and how it affects the formation of surface chemical gradients in the interior of these model pores. WCA mapping and ToF-SIMS imaging are used to visualize the plasma polymer deposit in the interior of the model pores and demonstrate that a strong dependence of the wettability gradient profile only exists up to a channel cross section of about 1 mm. XPS data allow us to calculate a deposition rate of plasma polymerized hexane (ppHex) at discrete positions on the surface and show that the deposition rate of ppHex is reduced by the presence of the mask up to a distance of 16 mm in advance of the channel opening. A strong dependence of the ppHex deposition rate on the cross-section of the channels is found within the first 2 mm in front of the pore opening. An estimation of the sheath thickness suggests that this effect can be attributed to the plasma sheath that perturbs the plasma in front of the pores. Plasma mass spectrometry allows us to identify the nature of the plasma species penetrating from the plasma through the pores and shows that no negatively charged ions are able to penetrate through the small channels. Neutral and positively charged species penetrate several millimeters down the channels and both species are therefore likely to contribute to the formation of the deposit on the sample. In addition, the formation of positively charged higher molecular mass hexane fragments is observed in the gas phase, demonstrating the likelihood of neutral-positive reactions in the plasma. 1. Introduction Gradients have emerged as powerful tools for high throughput screening in a number of applications such as tissue engineering,1 material development,2 and sensors.3 Surface gradients have been shown to be useful in tissue engineering, where uniform tissue growth within a porous 3D scaffold can be facilitated by even cell seeding.1 Gradients are especially suitable for rapidly screening the interaction of biomolecules with a range of material surfaces because the response to a wide range of different surface properties can be investigated on a single sample, thus increasing throughput and decreasing intersample variations and potential biological diversity between individual measurements.3 Surfaces with gradient properties have equally been applied to produce nanoparticle assemblies,4 achieve molecular self-assembly control,5 study surface orientations of liquid crystals,6,7 and induce uphill motion in water droplets.8 The preparation and analysis of gradient surfaces has developed into an important discipline in the field of biomedical materials development where surfaces with chemical gradients * To whom correspondence should be addressed. E-mail: [email protected]. Tel.: +44 (0)115 951 5119. Fax: +44 (0)115 951 5102. † University of Nottingham. ‡ University of Liverpool. § University of Glasgow. ⊥ Current address: University of Strathclyde, Institute of Pharmacy and Biomedical Sciences, The John Arbuthnott Building, 27 Taylor Street, Glasgow G4 0NR, U.K.

were shown to control relative cell adhesion, growth and motility as well as dictate the adsorption of proteins.2,9-11 While different approaches to preparing surface chemical gradients have been reported, plasma polymerization is so far the only method for the preparation of soft matter gradients that is not restricted to a specific substrate or functional group.2,12 This attribute, together with the possibility to coat complex 3D objects illustrated recently for tissue engineering scaffolds, makes plasma polymerization suitable for surface modification of a wide range of materials.1,13 When using plasma polymerization to modify the interior surface of porous objects, the availability of depositing species at any point within the item is limited by their diffusion through the pores.1 This results in the formation of gradients due to the decrease in thickness of the deposit from the outside of the object to its center. To quantify the deposit formation down the pores, here we report the first use of model pore channels consisting of two parts that can be separated to analyze the modified interior surface. Thus, the decrease of the plasma polymer deposition rate can be studied both as a function of distance from the opening and for a range of pore sections. Identification of the aforementioned depositing species is not simple; the mechanism of plasma polymerization has been debated for many years, moving from initial descriptions of deposition dominated by neutral plasma species14 to studies demonstrating the importance of positive ions based on plasma mass spectrometry15 to work highlighting the need to consider

10.1021/jp902137y CCC: $40.75  2009 American Chemical Society Published on Web 06/01/2009

8488

J. Phys. Chem. B, Vol. 113, No. 25, 2009

Figure 1. Schematic of the sample preparation procedure.

plasma phase negative polymeric ions16 and most recently the suggestion that high molecular mass ions in the gas phase could be noncovalently bound clusters that polymerize at the surface.17 Experiments under masks allowed the mechanism of plasma polymer deposition down small pores or through apertures that exclude the plasma to be described globally as diffusion controlled, with depletion of depositing species to the walls.18 However, the nature of the species contributing to the deposit from the plasma has not yet been unequivocally identified. To determine the importance of neutral, positive and negative plasma species in deposition within pores, here we employ mass spectrometry to analyze the ionic and neutral flux from a hexane plasma after passage through tubes of different lengths to mimic different points of the penetration of plasma species down pores. 2. Experimental Methods 2.1. Mask Preparation. A polyolefin plastomer (POP) block was prepared from POP granules (Affinity Polyolefin Plastomers, Dow Europe GmbH). A negative template of the channel structure was machined into a piece of brass. The brass template was sonicated in water and thoroughly washed with hexane three times to eliminate traces of silicon contamination from its surface. The positive structure was then transferred onto the POP block by hot embossing. 2.2. Gradient Formation. A custom built reactor made of borosilicate and stainless steel end plates sealed with Viton O-rings was used for plasma polymerization. The plasma was initiated via two capacitively coupled copper band ring electrodes connected to a radio frequency power supply (13.56 MHz). The monomer pressure was controlled via needle valves (BOC Edwards) and monitored with a pirani gauge (Kurt J. Lesker). A quartz crystal sensor (Sycon Instruments) inside the chamber was used to monitor deposition rate and film thickness in the plasma cloud. All depositions were carried out with a power of 20 W and under an operating pressure of 300 mTorr. Allylamine (99%) was obtained from Sigma Aldrich and hexane (HPLC grade) from Fischer Scientific. The monomers were subjected to at least one freeze-pump-thaw cycle before deposition. Microscope slides (VWR) were sonicated for 15 min in deionized water, washed with acetone, and cleaned in an oxygen plasma for 3 min. They were coated with ppAAm for 150 s at an average deposition rate of 0.40 nm s-1 on the quartz crystal sensor. The POP mask was placed on the sample and sealed with parafilm at three sides to prevent lateral diffusion of the plasma. The samples were placed in the reactor such that the openings of the channels faced the monomer feed (see Figure 1). Plasma polymerization of hexane was carried out for 142 s at an average deposition rate of 0.15 nm s-1 on the quartz crystal sensor. For the flow control experiment, a solid block of POP was placed on the sample with the ppAAm coated side facing in the opposite direction of the monomer feed (see Supporting Information, Figure 1). ppHex was deposited for 150 s at an

Zelzer et al. average deposition rate of 0.12 nm s-1 on the quartz crystal sensor. After each deposition, the samples were exposed to the monomer for an additional 3 min to saturate remaining reactive sites. All analysis was carried out within a week after sample preparation. 2.3. Surface Analysis. Water contact angles (WCA) were measured in 0.25 mm increments with pico-liter sized droplets using a DSA 1000 (Kru¨ss, Germany). The evolution of the drop shape was captured over 1 s after deposition of the drop on the surface. The WCA was calculated from the first stable image of the drop using a circular fit for the drop shape. WCA line profiles were obtained by averaging the WCA data for each channel. ToF-SIMS analysis was carried out on a SIMS IV time-offlight instrument (ION-TOF GmbH, Mu¨nster, Germany) equipped with a gallium liquid metal ion gun and a single stage reflectron analyzer. The instrument was operated typically at a primary ion energy of 15 kV, a pulsed target current of 1.3 pA and a post acceleration of 10 kV. Charges induced on the substrate surface by the positively charged ion beam were compensated with a flux of low energy electrons (20 eV). Large scale images were obtained by rastering the stage under the pulsed primary ion beam. All doses were kept below the static limit, with a maximum dose of 1012 ions per cm2 for both polarities combined. Acquisition of full raw data sets allowed for the retrospective construction of spectra from the imaged areas. Images are normalized to the total ion intensity. XPS spectra were acquired on a Kratos Axis Ultra (Kratos, UK) with a monochromated Al kR source (1486.6 eV) using an emission current of 15 mA and an anode potential of 12 kV. The instrument was used in fixed analyzer transmission mode (FAT) with a pass energy of 80 eV. The analysis area was defined by a magnetic immersion lens system with a 110 µm aperture. Charge neutralization was used and the takeoff angle for the photoelectron analyzer was 90°. 2.4. Plasma Mass Spectrometry. The penetration of plasma species through tubes of varying length was measured in a plasma reactor that was attached to a mass spectrometer (Hiden Analytical Ltd. EQP 300). The detailed experimental setup of the plasma reactor is described elsewhere.19 The hexane plasma was operated under the same conditions stated for the channel preparation (RF of 13.57 MHz at a power of 20 W), with the exception of the working pressure which was set to 100 mTorr due to limitations of the working pressure of the attached mass spectrometer. Glass tubes of different length (0 (no tube), 2, 4, 8, and 12 mm) with a square cross section of 0.5 mm × 0.5 mm were placed in front of the gate to the mass spectrometer, such that only species penetrating through the tubes would be detected. The glass tubes were mounted in a custom built holder (Supporting Information, Figure 2) that allowed each tube to be positioned in front of the mass spectrometer inlet without switching the plasma off. Neutral species were subjected to electron bombardment to ionize them. The electrons were accelerated to a potential of 50 V which is high enough to turn the neutral species into positive ions which were detected with the mass spectrometer while keeping fragmentation to a minimum. Even though the mass spectra were acquired from a hexane plasma operated at a working pressure that was lower than the one used for the preparation of the channels, comparison of mass spectra obtained at different pressure settings (data not shown) showed that the dependence of the spectra on the pressure is reasonably small. We therefore assume that the acquired mass spectra can be used to qualitatively explain the

Influence of the Sheath on Plasma Polymer Deposition

Figure 2. WCA map of the sample taken with picoliter sized droplets in 0.25 mm increments. White lines indicate the base-profile of the mask.

J. Phys. Chem. B, Vol. 113, No. 25, 2009 8489

Figure 3. WCA line profiles averaged over the width of the gradients. On the 0.25 mm channel, only one measurement could be taken: (red) 0.25 mm, (green) 0.5 mm, (blue) 1 mm, (orange) 2 mm, (black) 5 mm; error bars represent standard deviations.

processes involved in the plasma polymerization of hexane inside small pores under the previously reported conditions. 3. Results 3.1. Sample Preparation. To obtain geometrically welldefined model structures of controlled dimensions, a 1 cm thick polyolefin plastomer (POP) mask was prepared that contained five open channels with square cross sections that were 0.25, 0.5, 1, 2, and 5 mm in length. The underside of the mask was pressed against a glass slide precoated with plasma polymerized allylamine (ppAAm). This assembly was then exposed to a hexane plasma (Figure 1). The open end of the channels allowed penetration of depositing plasma species down the model pores to form a thickness gradient of plasma polymerized hexane (ppHex). The formation of the gradients was visualized by the complementary surface chemical analysis techniques of mapping picoliter sessile drop water contact angle (WCA) measurements, imaging time-of-flight secondary ion mass spectrometry (ToFSIMS), and X-ray photoelectron spectroscopy (XPS).20 3.2. Water Contact Angle Mapping. The wettability of the surface was mapped by water contact angle (WCA) measurements with picoliter sized droplets (Figure 2). The WCA map showed that the increase in wettability from hydrophobic ppHex to more hydrophilic ppAAm down the channels is steepest on the 0.25 mm cross sectional channel. Areas that were in direct contact with the mask mostly maintained the lower WCA of the ppAAm surface (∼66°), although some loss of pattern fidelity was observed for the larger channels. Figure 3 shows the WCA profiles of all five gradients obtained when averaging the data of each channel. The WCA transition from the hydrophobic to the hydrophilic end occurred over a shorter distance in the smallest model pore (transition over a length of 6 mm) compared to the larger channels. The dependence of the slope of the gradient profile (starting from x ) 0) on the size of the channels is illustrated in Figure 4. The slope of the WCA gradient increased rapidly from the 0.25 mm to the 1 mm channels and approached a constant value for larger cross sections. 3.3. ToF-SIMS Imaging. In the ToF-SIMS spectra, hydrocarbons (C-, CH-, CH2-, C2-, C2H-), nitrogen containing species (NH-, CN-, CHN-), oxygen (O-, OH-) and low intensities of fluorine and silicon containing ions (not present in the displayed range) were identified in the spectra and grouped together to form images of the separate components (Figure 5). The intensity of the hydrocarbon fragments was highest on the unmasked side of the sample (left side of image B in Figure

Figure 4. Change of the WCA on the gradients expressed as the slope of the WCA profile plotted against the size of the channel.

5b) and decreased inside the channels. The nitrogen containing fragments showed the opposite trend (image C). In both cases, the areas between the gradients that were covered by the mask could be clearly distinguished as dark (CH groups) and bright (N groups) regions on the images. The gradual change in CH and N group intensities was shorter in the small channels. Image F shows a convolution of the hydrocarbon (red) and nitrogen groups (blue) where the counter-gradients of CH and N groups are displayed. Less pronounced gradient transitions were also observed for the oxygen containing fragments (image A in Figure 5b) that showed an increasing intensity when moving along the gradients. The Si-containing ions observed were attributed to polydimethylsiloxane (PDMS) contamination; their distribution was largely uniform across the whole sample. Fluorine ions were only observed on those areas of the sample that have been in contact with the mask and can therefore be attributed to contaminations from the POP block. 3.4. XPS Analysis. The chemical composition of the surface was analyzed by small area XPS to calculate the ppHex thickness profile of the gradients (spectra not shown). Carbon, oxygen, and nitrogen but no silicon were detected on the surface. The oxygen contentsincorporated in the plasma polymer due to postoxidation21,22 swas low and constant (