Atmospheric Pressure Plasma Treatment of Polymers: Relevance to Adhesion

By Michael Thomas and K.L. Mittal

The Atmospheric Pressure Plasma (APP) treatment for polymer surface modification has attracted much attention recently, owing to its advantages over other techniques and its ability to improve adhesion without tampering with polymer's bulk properties. Focusing on the utility of APP treatment for enhancing polymer adhesion, this book covers the latest development in this important and enabling technology, providing profound insights from many top researchers on the design and functions of various types of reactors, as well as current and potential applications of APP treatment.

• Language: English
• Publisher: Wiley
• Release date: Jun 19, 2013
• ISBN: 9781118747513

Book preview

PART 1

FUNDAMENTAL ASPECTS

Chapter 1

Combinatorial Plasma-based Surface Modification of Polymers by Means of Plasma Printing with Gas-Carrying Plasma Stamps at Ambient Pressure

Alena Hinze¹, Andrew Marchesseault², Stephanus Büttgenbach², Michael Thomas³ and Claus-Peter Klages¹,³

¹Technische Universität Braunschweig, Institut für Oberflächentechnik (IOT), Braunschweig, Germany

²Technische Universität Braunschweig, Institut für Mikrotechnik (IMT), Braunschweig, Germany

³Fraunhofer Institute for Surface Engineering and Thin Films IST, Braunschweig, Germany

Abstract

In this work a new method of achieving combinatorial area-selective modification of polymer surfaces is presented, utilizing atmospheric-pressure plasma printing with novel gas permeable electrodes. In these plasma stamps a microporous gas-carrying layer provides exchange of gaseous species from the gas stream to the individual microcavity discharges. Additionally, the electrodes can be fed with two (or more) different gases from spatially separate locations, allowing the generation of spot arrays with controlled gradients of physicochemical surface properties. Plasma-printed gradient surfaces can be used for combinatorial studies, for example in biomedical or polymer electronic research. In combination with spatially resolved surface characterization methods, the investigation of plasma-surface interaction processes can be significantly simplified. In the present contribution, gradient spot arrays were applied to optimize gas composition and functionalization parameters to provide optimal nucleation and growth of an electroless metal coating on a polymeric substrate. Locally plasma-modified surfaces were quantitatively characterized applying chemical derivatization (CD) followed by FTIR-ATR or SEM-EDX analyses in order to determine the area densities and spatial distributions of functional groups which are reactive towards the derivatization reagents used. Two chemical derivatization techniques were utilized: gas-phase derivatization (i) with 4-(trifluoromethyl)benzaldehyde (TFBA), forming a stable Schiff base with primary – but not secondary- amino groups, and (ii) with 4-(trifluoromethyl)phenyl isothiocyanate (TFMPITC) which is able to react with both primary and secondary amino groups forming thioureas, but – under the conditions used – not hydroxyl groups. It was, however, recently pointed out by us that other nitrogen-bearing functional groups such as imines can be captured by these methods as well.

Keywords: Dielectric barrier discharges, plasma printing, microplasmas, porous plasma stamps, polymer surface modification, gradient arrays, combinatorial plasma chemistry

1.1 Introduction

The term plasma printing stands for patterned surface modification or plasma-enhanced film deposition using ambient-pressure microplasmas enclosed in sub-millimeter sized cavities [1]. In early investigations, ceramic plates with laser- or mechanically drilled cylindrical through-holes covered by a fine metal mesh were used in order to allow diffusive gas exchange between the cavities and ambient. The mesh simultaneously served as one of two discharge electrodes, providing the electric field necessary to ignite a barrier discharge within the cavity. Using such an arrangement, the process gas can be transported by a stagnant flow and diffuse through the mesh into the cavities below it, enabling surface treatment with larger amounts of gas than available in the enclosed cavity volume. Different kinds of thin films with thicknesses up to several 100 nm have been deposited with arrangements of this kind [2, 3].

Producing more complicated patterns or larger arrays of regular vias in ceramic plates by laser-based or mechanical methods, however, is not trivial. Using free-standing insulator sheets for the definition of the plasma-printed areas, it is generally impossible to generate patterns in which the non-treated areas are not connected. In addition, a good mechanical contact between the plasma stamp and the substrate to be treated cannot be easily guaranteed because the application of uniformly distributed mechanical forces interferes with the provision of a stagnant gas flow over an area of several square centimeters.

For these reasons, recent work on plasma printing has focused on plasma stamps with closed cavities, produced by photolithographic techniques or by electromagnetic engraving [4] which were used for patterned plasma nitrogenation¹ or plasma oxidation of polymer surfaces with lateral dimensions on the order of 100 μm or lower.

In order to make the patterned deposition of thicker coatings using plasma printing feasible, new solutions are required for the assembly of dimensionally stable plasma stamps that provide an exchange of gaseous species with the discharge in the cavity. Interesting opportunities are afforded by utilizing recent developments in the field of porous metallic materials, such as components with high permeability and porosity, which can be obtained from the sintering of metal fibers [6].

The principle of plasma stamps with a porous gas-carrying layer is illustrated in Figure 1.1. Compared with closed versions of plasma stamps the new design utilizing a microporous gas-carrying layer as an electrode offers a number of advantages:

Figure 1.1 Patterned surface modification of an insulating substrate (white layer) using a plasma stamp with a porous metal plate as a gas-carrying electrode (checkered layer). The ground electrode is an ITO-covered (thin black) glass substrate (dark grey). Convective gas transport through the porous metal plate is indicated by horizontal arrows, the predominantly diffusive gas exchange within a single cavity (vertical hatching) is defined by a polydimethylsiloxane (PDMS) mask layer (light grey) shown as vertical arrows. The top electrode (thick black) is powered by high voltage source. The plasma stamp is the part enclosed by the dashed rectangle.

If surfaces shall not only be modified, like in the present paper, but when thin films beyond a few nm thickness are to be deposited, gas-carrying plasma stamps are a big advantage because virtually unlimited gas volumes can be fed into the micro-cavities and can be used for film formation.

The cavities formed by the substrate and the stamp can be fed with the gas quite rapidly and very spa ringly. Oxygen traces in the cavities can be quickly displaced while getting around the necessity to provide an oxygen-free environment. This point is very important if polymer surfaces are to be plasma-nitrogenated because oxygen molecules compete with intermediate radical centers in this process [7].

During the plasma treatment or coating of a surface, typically lasting a few seconds, a time-independent gas composition will be guaranteed due to the diffusive gas exchange between the microplasma in the cavities and the gas stream in the gas-carrying layer. Products of the process will be continuously swept away and a redeposition of plasma polymers will be suppressed.

Last, but not least, plasma stamps with a gas-carrying layer based on a highly porous plate make new methods of combinatorial investigations of plasma-chemical surface modification processes possible and are very helpful for finding optimal process parameters.

Combinatorial methods in surface science and technology can be used to prepare gradient surfaces, i.e., surfaces with physicochemical property gradients which have recently received a lot of interest [8–10]. Gradient surfaces have a discrete or continuous spatial variation in physicochemical properties such as surface free energy, chemical composition and functional group densities or charge densities. A key benefit of gradient surfaces is that a small number of samples can be used to investigate the effect of variation of preparation parameters on many surface properties. The use of gradients significantly reduces time and improves the efficiency of R&D work. Gradient surfaces have been successfully used for combinatorial high-throughput studies in the search for new catalysts, semiconductors or superconductors and in biomedical or biomaterials research.

While combinatorial methods are well-established in chemistry and biochemistry systematic studies in plasma-based surface science are relatively rare and only in the recent years have several papers appeared on this topic, such as refs. [11, 12]. A method in which the movement of a substrate table was synchronized with a change in gas mixture being fed through an atmospheric-pressure plasma processor was published in 2004 [13]. The advantage of a method based on porous metal electrodes is that it can be made in a planar fashion and that the library of thin films compositions is generated in an area of only a few square centimetres, well-suited to surface analyses with physical methods, see below. For virtually every application in which the determination of optimum conditions for film deposition or surface modification by a plasma-based surface-technological process is necessary, combinatorial techniques can be used to drastically diminish the experimental effort.

An example of our current interest is the optimization of gas composition and deposition parameters to achieve optimal nucleation, growth and adhesion of electroless metal coatings on polymeric substrates [4], though many other applications are imaginable. Thus, the investigation of plasma-surface interaction processes in combination with spatially resolved surface characterization methods can be substantially simplified.

1.2 Experimental

1.2.1 Porous Plasma Stamp Design and Fabrication

The heart of the novel plasma stamp is a 36 × 36 × 5 mm³ porous metal plate, manufactured by the sintering of Cr-Ni steel fibers (IFAM, Dresden, Germany) (see Figure 1.2). Steel fibers with a diameter of 27 μm were utilized in order to enable high gas permeability, with an open pore volume of about 84%, and still have a quasi-homogeneous global electric field distribution within the cavities.

Figure 1.2 SEM image of the porous metal plate provided by IFAM (Dresden, Germany).

As an electrode enclosure, a polycarbonate fixture was manufactured (see Figure 1.3) and equipped with quick connecting hose fittings screwed into threaded inlet and outlet bores. The fixture was designed with a flat aluminum backing plate to allow an electrical connection from the sintered porous electrode to ground, and to allow efficient heat transfer for improved cooling during plasma operation. Before the porous electrode was placed into the polycarbonate enclosure, the backside of the gas-carrying sintered porous plate as well as the lateral sides had been sealed with an aluminum foil in order to prevent gas leakage during plasma treatment.

Figure 1.3 Polycarbonate fixture with inserted porous metal plate, flow distributors, threaded connectors and aluminum backing plate.

In order to define the cavities a microstructured polydimethylsiloxane (PDMS) mask layer carrying an array of through-holes was applied to the metal porous plate. Several methods for producing such layers from PDMS are reported in literature [14]. Both, a compression method and spin coating are feasible. The compression method studied initially was a variant of work in double-sided relief molding [15], where a planar counter weight was pressed against the microfabricated mold. This method was successfully used to fabricate circular structures down to 500 μm in diameter and at an offset of 1250 μm, center to center. If the force applied was too low, however, it was found that a film of PDMS remained on the top of the structures, therefore producing closed cavities. Due to surface tension effects, the amount of force had to be increased with more closely packed structures. At higher structure densities than that of the previously mentioned limit, the force became so large that the risk of destroying the mold was too high. For that reason it was decided to apply a spin coating technique for the PDMS layers.

Spin coating of PDMS is a very common method in literature for producing extremely thin membranes, including through-holes. Structures with diameters as low as 125 μm and offsets of 450 μm have been successfully fabricated. The thickness of these layers is currently the limiting factor, as the handling of structures less than 200 μm in thickness often leads to tearing. Layers, therefore, are fabricated in thicknesses of 240 μm or more. In order to increase the intensity of a signal during ATR-FTIR measurements and make quantitative characterization of plasma-printed polymer surfaces easy, PDMS film with cavities of relatively large diameters (500 μm) centered on a triangular lattice with 750 μm lattice constant (center-to-center offset) and 240 μm heights were fabricated (see Figure 1.4) and used for the generation of locally modified gradient arrays. These masks yielded an area fraction q = 38.0 ± 1.8% during the FTIR-ATR measurements of the average functional groups densities.

Figure 1.4 Planar (a) and cross-sectional views (b) of a 240 μm thick PDMS layer with 500 μm cavities and 750 μm offset.

Figure 1.5 shows how the metal porous plate, a PDMS layer carrying a hexagonal array of through-holes, and a polymer foil, serving as the substrate to be plasma-printed, are arranged. As a counter electrode, a 500 μm thick glass plate covered with transparent indium tin oxide (ITO) is applied on top of this arrangement, in order to enable visual inspection of the microdischarges in the cavities from above. During the initial experiments it was discovered that the uniformity of plasma generation within individual cavities could be improved considerably by the insertion of a fine steel wire mesh between the porous plate and the PDMS layer, as described below.

Figure 1.5 Preparation for plasma printing: a porous plasma stamp covered from above with a polymer foil.

1.2.2 Plasma Printing

Plasma-printed polymer surfaces with controlled gradients of nitrogen-bearing functional group densities were prepared using an experimental arrangement as shown schematically in Figure 1.6. The plasma printing setup described in detail elsewhere [16] was modified slightly: The assembly is based on the principle of two vacuum chucks compressed together to align and compress a polymer substrate with the porous plasma stamp, providing constant flow of process gases along the joined cavities with well-controlled spatial extension of the plasma. In the new assembly, the counter ITO electrode is vacuum-fixed to the upper chuck made of 15 mm thick acrylic glass. Transparency of both the ITO electrode and acrylic chuck permits clear observation of plasma ignition in the cavities. A flexible acrylic polymer was chosen in order to avoid destruction of the ITO-coated glass electrode when compressed together with the porous plasma stamp. The bottom vacuum chuck, a grounded aluminium table, is driven by an electric mini-slide, so that the porous plasma stamp on it, covered with a polymer substrate, can be rapidly raised against the counter ITO electrode and compressed with it at an adjustable force up to 196 N. The gas in the plasma stamp is fed by two gas hoses from spatially separate locations. Additionally, gas purification system with commercial Oxysorb® and Hydrosorb® cartridges (Messer Griesheim. Ltd, Germany) was applied to minimize O2, H2O and CO2 contaminations for each process gas.

Figure 1.6 Scheme of a plasma printing assembly with a porous plasma stamp in detail.

A stainless steel square wire mesh (25 μm bar width, 25 μm hole width, PACO Paul Ltd. & Co, Germany) was placed between the PDMS film and the porous metal in order to enhance the planarity of the gas permeable electrode surface, leading to an improved homogeneity of plasma treatment within spots on polymers surfaces. Preliminary visual inspection of distribution of plasma light emission within the cavities during modification in N2 and N2 + 4% H2 process gases without this mesh had indicated strong non-homogeneities. This was caused by relatively thick electrode fibers (27 μm) arranged under the PDMS masking vias (500 μm) and the tendency to generate higher electric fields at the sharp edges of the fibers. A positive effect of using the steel wire mesh to enhance uniformity of modification within treated plasma spots was also demonstrated by means of fluorescence intensity distributions investigation after plasma modification and labeling with fluorescamine (Fluram™, Fluka). The applied routine labeling technique, which used fluorescamine to mark primary amines introduced locally onto the polymer surface is described elsewhere [16]. Selectivity of labeling with fluorescamine of primary amino groups for fluorescence measurements is currently under investigation.

Gradient polymer surfaces with various functional group densities were plasma-printed with 20 s of plasma exposure at 75–90 N of contact force in virtually oxygen-free N2 (inlet 1) and N2 + 4% H2 mixtures (inlet 2). In order to displace trace oxygen from the cavities, the plasma stamp was flushed with process gases for 3 min before plasma operation. Process gas flows of 125 sccm were employed for both gas streams. Biaxially oriented polypropylene foil of 75 μm thickness (BOPP, Goodfellow Ltd, Germany) was utilized as the substrate material. The polymer substrates were pre-cleaned by washing in isopropanol, in acetone and were dried finally in a stream of pure N2. The plasma printing assembly was powered by a mid-frequency generator 7010 R and a high-voltage transformer AT 7010 R (Softal Electronic Ltd, Germany) operated with a sinusoidal wave signal with amplitude of 5.4 kV at 23 kHz. Due to the tendency of the PDMS masking layer to form plasma-etched products, PDMS cavity walls were passivated by plasma oxidation in air (4.7 kV, 23 kHz) with 2 min of plasma exposure, followed by plasma ignition in the cavities in a controlled atmosphere of process gases.

1.2.3 Chemical Derivatization of Functional Groups

Aside from amino groups, many other functional groups are commonly incorporated into polymer surfaces exposed to a N-containing gas discharge. For quantitative characterization of plasma-modified gradient surfaces, the area densities and spatial distributions of functional groups were evaluated, utilizing two different gas-phase derivatization techniques. In order to selectively label primary amino groups, the well-established gas-phase derivatization with 4-(trifluoromethyl)benzaldehyde (TFBA, Sigma-Aldrich Ltd, Germany) was used, resulting in the formation of surface-bound trifluoromethylbenzaldimine groups. Gas phase derivatization with 4-(trifluoromethyl)phenyl isothiocyanate (TFMPITC, Sigma-Aldrich Ltd, Germany) was performed in order to capture both primary and secondary amino groups, while avoiding capturing of hydroxyl groups which might also be formed on the surface. Isothiocyanates are known to react with primary and secondary amines to form thioureas. 3,5-bis(trifluoromethyl)phenyl isothiocyanate derivatization of several amino-bearing model surfaces was already studied by Graf et al. [17] using XPS and NEXAFS. According to the authors, not all amino groups were necessarily captured by this method, depending on the type of thin film studied. However, our studies show that imines are able to react with both derivatization reagents as well [5]. Our so far unpublished results obtained with TFMPITC have shown that

in model reactions with low-molecular primary amines in aliphatic hydrocarbons, monitored by FTIR-ATR, 4-(trifluoromethyl)phenyl isothiocyanate reacted faster than the corresponding aldehyde TFBA,

quantitative FTIR-ATR analysis carried out on plasma-modified polymer surfaces after gas phase derivatization with the isothiocyanate usually yielded larger amounts of CF3 groups than after labeling with aldehyde, indicating a significant amount of secondary amines,

CD-FTIR-ATR performed with both aldehyde and isothiocyanate on pulsed plasma polymerized surfaces from aminopropyltrimethoxysilane (APTMS), yielded, within experimental errors, similar results, as is to be expected for near-complete retention of amino groups,

appreciably lower amounts of CF3 groups were obtained in experiments on pulsed plasma polymerized surfaces from APTMS in which CD-XPS was used to determine the fluorine content after derivatization with the isothiocyanate, compared with CD-ATR. This was normally not observed with the aldehyde TFBA, and

almost no CF3 groups were detected by means of CD-SEM-EDX on pulsed plasma polymerized spots from APTMS derivatized with TFMPITC.

These results have encouraged the combination of aldehyde and isothiocyanate gas-phase derivatizations, combined with FTIR-ATR analysis in order to determine separately the densities of TFBA- and TFMPITC-reactive functional groups, ρTFBA and ρTFMPITC, obtained by plasma modification.

In order to evaluate the remnant stable plasma-introduced functional groups after exposure in strong solvents, as well as to avoid interference of loosely bound low-molecular weight components on the polymer surface, freshly modified gradient spots arrays were first immersed in acetone for 5 min and then dried with a nitrogen stream. Subsequently, the specimens were fixed on a special holder in a closed 250 ml glass vessel and exposed to either vapors of 0.5 ml TFBA in order to selectively label TFBA-reactive groups or to vapors of 0.5 g TFMPITC in order to label TFMPITC-reactive groups. The derivatization reactions were run for 4 h under a protective Ar atmosphere. The polymer specimens were placed under vacuum overnight at 10−3 mbar in order to remove physisorbed reagent molecules from their surfaces.

1.2.4 FTIR and EDX Analyses

Quantitative chemical analysis of plasma-modified and derivatized gradient polymer surfaces with respect to the area densities of nitrogen-bearing functionalities was carried out utilizing (i) Fourier transform infrared spectroscopy in the attenuated total reflectance mode (FTIR-ATR) and (ii) energy dispersive X-ray analysis in a scanning electron microscope (SEM-EDX).

The plasma-functionalized and derivatized depth of a polymer surface is typically up to 10 nm for dielectric barrier discharge (DBD) treatment. Both analytical methods, namely FTIR-ATR and SEM-EDX, used in this study have a probing depth much larger than the plasma-functionalized depth and can therefore be exploited to reveal the density of the species. Moreover, electron probe X-ray spectroscopy can also be used to determine the spatial distribution of functional groups with a sub-μm resolution.

To record FTIR-ATR spectra a Nicolet 5700 FTIR instrument was used which is equipped with an MCT detector and a DuraSamplIR single reflection 45° diamond ATR crystal using unpolarized light and a spectral resolution of 1 cm−1. The strongest signal at 1323 to 1325 cm−1 assigned to the C-CF3 stretching vibrational band of surface-bound TFBA- or TFMPITC-derivatized moieties was used to determine the area densities of functional groups on gradient polymer surfaces. The sampling depth of the ATR method is roughly 1 μm. The area density evaluation of TFBA- or TFMPITC-reactive groups from CD-FTIR-ATR measurements, described in detail elsewhere [18], is performed by comparing ATR spectra of plasma-printed derivatized polymer substrates with suitable reference solutions, assuming that molar absorption coefficients for the characteristic vibrations in the polymer and the reference solution are equal.

The area density ρ of derivatized groups within the ultrathin locally functionalized surface region is evaluated according to Eq. (1.1):

(1.1) equation

where APq is the hypothetical absorbance of a uniformly derivatized surface layer covering the complete sampled area with the same group density as the (average) density on the microspots; c is the known concentration of CF3 groups in the reference solution; dp is the penetration depth; AR is the absorbance of the reference solution. The microspots printed on the polymer surface cover only a part of the ATR crystal. This results in an area fraction q defined as the ratio between the spot area of 0.7105 ± 0.0331 mm² under the ATR crystal and the area of ATR crystal of 1.87 mm². APq and reflectivity RPq are calculated from the measured absorbance of the polymer AP according to Eq. (1.2):

(1.2) equation

In order to obtain quantitative information about area densities and spatial distribution of TFBA- and TFMPITC-reactive groups on BOPP foil with a gradient plasma-modified array, SEM-EDX analyses after the chemical derivatization procedure with TFBA and TFMPITC were carried out using a scanning electron microscope system (Leo 1530, Oxford Instruments EDX Microanalysis System, Ge detector). Measurements of fluorine introduced onto the surface after derivatization were performed at 1.5 keV primary electron energy. The depth of the analysis is given by the ultimate depth of X-ray emission (de). X-ray absorption in these investigations was not too strong being almost identical to the maximum depth of the X-ray generation (dmax). This value can be calculated from Castaing’s formula [20]:

(1.3)

equation

where ρ [g/cm³], A and Z are the density, mean atomic weight and the mean atomic number, resp., of the investigated material; EC [keV] is the critical excitation energy of the X-ray line. At a primary electron energy of E0 = 1.5 keV the maximum X-ray emission depth de ≈ dmax for F Kα line (EC = 0.67 keV) in a typical polymer is about 100 nm.

The evaluation of the CD-SEM-EDX measurements is described in detail elsewhere [16, 19]. Poly(tetrafluoroethylene) (PTFE, Goodfellow Ltd, Germany) was chosen as a reference sample for the measurements. The reference sample and the polymer gradient array were coated with a 10 ± 1.5 nm layer of evaporated carbon deposited with an MED 020 high vacuum coating system (Bal-Tec, now Balzers AG, Liechtenstein) at a defined distance of d = 7.5 cm. This was done in order to make the polymer surfaces electrically conductive and simultaneously diminish fluorine loss by radiation damage during the analysis. Carbon coating thickness was additionally analyzed for every investigated array by use of spectroscopic ellipsometry measurements utilizing SE850DUV instrument (Sentech Instruments GmbH, Germany). These investigations were performed on C-coated Si wafers deposited with a combination of the reference sample and the polymer gradient array. Spectral simulation (STRATAgem program, SAMx, France) was employed in order to derive the relation between measured carbon and fluorine peak areas and the density of fluorine atoms. The density of fluorine atoms is three times the original TFBA- or TFMPITC-reactive groups density [16]. To speed up time-demanding CD-SEM-EDX measurements taken from every 3rd locally defined plasma-functionalized spot, a spatial resolution of 50 μm was chosen, although a much higher resolution on the order of 1 μm can be achieved with this analysis in principle.

1.2.5 Electroless Metallization

The metallization of gradient spots arrays was carried out according to a procedure described in detail elsewhere [21]. At first, the plasma-modified specimens were immersed for 5 min in a PdCl2 solution at room temperature and subsequently in an aqueous sodium hypophosphite solution at 70°C in order to form a palladium catalyst layer (sensitization process). Plasma-activated N-containing groups are known to promote the chemisorption of palladium [22]. Reductive deposition of copper on the sensitized polymer surface was performed at 25°C with a deposition time of 2 min from a non-commercial bath containing copper sulphate, sodium hydroxide, potassium sodium tartrate and formaldehyde.

1.2.6 Numerical Simulation of Concentration Distributions

As previously mentioned in Section 1.2.1, Computational Fluid Dynamic (CFD) simulations were conducted using software Fluent™ in order to optimize the velocity and concentration profiles of gases in the porous electrode inserted into an acrylic enclosure. In this way, a general insight into the possible resulting plasma-printed gradients was obtained.

Since the permeability of the electrode was not given by the supplier, it was initially estimated and later verified experimentally to be 10−10 m². Using varying inlet conditions, including flow rate and gas composition, the velocity profiles were compared. After analyzing these results, it was determined that inlets and outlets leading directly into the porous region hindered the desirable distribution of the flow. It was, therefore, decided to use a flow spreader and nozzle at the inlet and outlet in order to develop a homogeneous flow profile.

After the optimal velocity profile was obtained, the end goal of mixing two process gases in order to obtain a diffusion gradient could be started. For this purpose, two flow spreading inlets were used with a single outlet. Two different feed gases were then utilized and a diffusion model based on gas kinetic theory was applied. After insertion of the flow spreaders, a very uniform velocity distribution (not shown here) and a symmetric funnel-shaped hydrogen concentration distribution could be obtained (Figure 1.7).

Figure 1.7 The numbers correspond to H2 concentration distribution (mol%) obtained within the porous metal plate as calculated by Fluent™. N2 and N2 + H2 gas flows enter the porous plate homogeneously distributed over the upper and the lower halves of the left edge of the plate, resp.

1.3 Results and Discussion

Gradient polymer surfaces were successfully fabricated at atmospheric pressure using porous plasma stamps and a short contact with microcavity discharges in N-containing gases with a concentration gradient of H2 in the plasma. Due to the fact that plasma-modified thin film compositions were generated over a relatively small area, 36 × 36 mm², the samples were well-suited to carry out surface analyses with quantitative FTIR-ATR and SEM-EDX methods. Combinatorial studies on plasma-printed gradient arrays were conducted in order to discover the influence of hydrogen concentration c in nitrogen (0 < c < 4%) on functional groups densities and their subsequent metallization behaviour. According to the standard metallization procedure, plasma-modified polymer samples have to be immersed into various polar solutions for a few minutes several times. Therefore, the BOPP gradient spot arrays were firstly exposed to acetone for 5 min followed by derivatization with TFBA or TFMPITC in order to capture only stably introduced functional groups.

To determine the area density distribution of plasma-generated TFBA- or TFMPITC-reactive groups on gradient surfaces, the samples were characterized by means of CD-FTIR-ATR at six different locations (see Figure 1.8), corresponding to six different hydrogen concentrations on the spots as shown in Figure 1.7. The average values of TFBA- and TFMPITC-reactive groups densities within one location were calculated from three CD-FTIR-ATR measurements, which covered a large number of spots. The average value of functional groups densities that were introduced area-selectively were determined to lie between <1.2 up to 4.3 (ρTFBA) / nm² for moieties labeled with TFBA and from 2.8 up to 7.3 (ρTFMPITC) / nm² for functionalities derivatized with TFMPITC (see Figure 1.8). Applying both derivatization techniques, maximum average values of TFBA- and TFMPITC-reactive groups densities were determined at the locations with maximum hydrogen concentration, while minimum average values were measured on the areas with minimum hydrogen concentration. The results of quantitative CD-FTIR-ATR analysis demonstrate that a sufficient amount of functionalities remain on the polymer surface to be measured after 5 min of exposure of the gradient spot array to acetone. The quantification limit of FTIR-ATR analysis with the area fraction of q = 38.0 ± 1.8% during the measurements of the average functional groups densities by local modification of the polymer surface was about 1.2 functionalities per nm² The funnel-shaped concentration distribution of calculated functional groups densities corresponds well with the distribution of hydrogen concentration taken from the simulation of the diffusion profiles within plasma printing zone.

Figure 1.8 Average densities of TFBA- and TFMPITC-reactive groups determined with CD-FTIR-ATR at the indicated locations. Results are represented as functional group densities: ρTFBA/ρTFMPITC. Derivatization was carried out after 5 min of exposure of the substrate to acetone. Spatial distributions of TFMPITC-reactive groups densities ρTFMPITC within individual spots located in the upper left and the lower left regions were also calculated from CD-SEM-EDX measurements and are shown in Figure 1.9.

Figure 1.9 Spatial distributions of TFMPITC-reactive groups densities calculated from CD-SEM-EDX measurements across the diameter of plasma-printed 500 μm spots located in the upper left (shown as a continuous black line) and the lower left (shown as a dotted grey line) regions demonstrated in Figure 1.8. Derivatization with TFMPITC was carried out after 5 min exposure of the substrate to acetone.

After chemical derivatization of the gradient array with TFMPITC, several plasma-printed spots were chosen from the upper left and lower left regions in Figure 1.8, which corresponded to the minimum and maximum hydrogen concentration, resp. and were analyzed by means of CD-SEM-EDX. This was done in order to investigate the uniformity of modification within selected single spots. According to the CD-SEM-EDX results, a relatively uniform sum density distribution of TFMPITC-reactive groups on selected spots was observed (see Figure 1.9). The average TFMPITC-reactive groups densities, ρTFMPITC, measured with CD-SEM-EDX on these regions were calculated to be 2.0 ± 0.2 / nm² and 3.6 ± 0.4 / nm², demonstrating lower values compared with the TFMPITC-reactive groups densities obtained from CD-FTIR-ATR measurements. This effect is most likely due to fluorine loss by radiation damage during the analysis. CD-SEM-EDX measurements were not performed on TFBA-derivatized gradient arrays in this work. However, our unpublished results show that in the case of labeling with TFMPITC, fluorine loss is always higher compared with the values detected from the surface of TFBA-derivatized samples. This tendency was also observed from the CD-SEM-EDX results performed with both aldehyde and isothiocyanate on pulsed plasma polymerized surfaces from APTMS. Fluorine density measured with SEM-EDX on TFBA- and TFMPITC-derivatized surfaces was not the same, as is to be expected for near-complete retention of amino groups. Almost no fluorine signal was detected on TFMPITC-labeled specimens. Lower stability of derivatization reagent TFMPITC compared to TFBA during CD-EDX and CD-XPS analyses has been observed. However, average amino group densities calculated from CD-FTIR-ATR measurements performed with both aldehyde and isothiocyanate on pulses plasma polymerized surfaces from APTMS yielded, within experimental errors, similar results.

Uniformity of plasma modification within plasma-treated areas appeared to be much better than expected when considering the non-uniformity of plasma emission intensity distribution in the visible region of the plasma. Maximum intensities appeared at the center and the periphery of the individual cavities (see Figure 1.10).

Figure 1.10 Non-uniformity of plasma intensity distribution with maximum intensity at the centre and at the boundary of the spots during plasma printing with N2 (125 sccm) and N2 + 4% H2 (125 sccm) plasma gases.

A similar observation was already made in an earlier ToF-SIMS study in which the distribution of visible radiation from cavity discharges in N2 + 4% H2 was compared with the distribution of various secondary ions originating from functional groups generated simultaneously on a BOPP surface [23]: A virtually uniform distribution of −CH2−NH2 moieties appeared on the surface, recognizable by the CH4N+ secondary ion, which was monitored in spite of the fact that intentionally non-uniform discharge was obtained by utilizing relatively small overvoltages (difference between applied voltage and ignition voltage). These results can be interpreted by the important roles presumably