Plasma-enhanced chemical vapour-deposited films of di(ethylene glycol) dimethyl ether were analysed by a combination of X-ray photoelectron spectroscopy, atomic force microscopy, quartz crystal microbalance with dissipation monitoring (QCM-D), X-ray and neutron reflectometry (NR). The combination of these techniques enabled a systematic study of the impact of plasma deposition conditions upon resulting film chemistry (empirical formula), mass densities, structure and water solvation, which has been correlated with the films' efficacy against protein fouling. All films were shown to contain substantially less hydrogen than the original monomer and absorb a vast amount of water, which correlated with their mass density profiles. A proportion of the plasma polymer hydrogen atoms were shown to be exchangeable, while QCM-D measurements were inaccurate in detecting associated water in lower power films that contained loosely bound material. The higher protein resistance of the films deposited at a low load power was attributed to its greater chemical and structural similarity to that of poly(ethylene glycol) graft surfaces. These studies demonstrate the utility of using X-ray and NR analysis techniques in furthering the understanding of the chemistry of these films and their interaction with water and proteins.
The study of plasma polymer coatings that could potentially be used in biomedical devices is an area of increasing research interest [1–5]. One of the most common classes of thin film treatments employed in biomaterial devices is that of ‘low-fouling’ or ‘stealth’ coatings . These coatings need to be able to resist or inhibit protein adsorption within the body. In the biomedical materials field, the most common surface coating used to render a material resistant against protein fouling is through the use of poly(ethylene oxide) (PEO), also known as poly(ethylene glycol) (PEG) . PEG polymers have a number of properties that have been implicated in their low-fouling nature. ‘Steric repulsion’ and the effect of the ‘water barrier’ resulting from the structuring of water in the near environment of the PEG chains are two of the most commonly described theories in the literature [8–10]. To date, the largest effort in the plasma polymer field to produce PEG-like films has been through the use of monomers containing ethylene oxide units, typically glycol diethers, which are also commonly known as the ‘glyme’ family of monomers . Within this class of molecules, the preferred thin-film deposition technique used has involved the use of pulsed plasmas with the monomer tetraethylene glycol dimethyl ether or ‘tetraglyme’ [11–13]. Lopez et al.  first described the use of ‘glyme’ monomers to deposit low-protein-fouling plasma polymer surfaces. It is believed that the molecular structure of PEG-like plasma polymer films consists of randomly cross-linked methyl-terminated ethylene oxide chains. By controlling the plasma power during deposition, the ether content of PEG plasma polymer films can be controlled to an extent. Previous studies have shown that plasma polymer PEG-like films produced with monomers consisting of two or more ethylene oxide units can exhibit low-fouling properties, while monomers with one ethylene oxide unit do not . Recent studies by Johnston et al.  have found that films produced from higher molecular weight precursors retain longer fragments of intact monomer. Protein adsorption results suggested that, in general, protein resistance improves as the number of ethylene oxide units in the monomer precursor increases.
During the plasma deposition process, an activated monomer introduced in the gas phase under vacuum undergoes fragmentation, excitation and ionization. These fragments rearrange and react to form a cross-linked polymer matrix. The physico-chemical properties of a plasma polymer deposited from a specific monomer may be quite different from those of a conventional polymer. For example, the hydrogen content of plasma polymer films is low in comparison with the polymer of the corresponding monomer owing to considerable cross-linking . It is therefore highly unlikely that low-fouling plasma polymer films produced via the plasma polymerization of ‘glyme’ monomers will exactly reproduce the polymer surface chemistry generated from the more commonly used PEG polymer graft surfaces [9,17–19]. Plasma polymer chemistries generated from a particular monomer may vary substantially depending on the operational conditions used (reactor geometry, monomer flow rate, pressure, mode and the strength of power delivery and frequency) .
A number of surface characterization techniques have been used to investigate the physical and chemical properties of PEG-like plasma polymer films to aid in elucidating their surface properties. Techniques such as X-ray photoelectron spectroscopy (XPS), Fourier transform infrared spectroscopy (FTIR), Raman spectroscopy and mass spectrometry methods (e.g. time of flight secondary ion mass spectrometry) are typically used to deduce the amount of residual ether groups in these plasma-polymerized surfaces. However, there have been no reports on the full chemical composition (including hydrogen) of these ‘glyme’ plasma polymer films and their solvation in water to the best of our knowledge. Conventional surface characterization techniques that are commonly employed in the field do not quantify the amount of hydrogen remaining in plasma polymer films after plasma polymer deposition. Reflectometry techniques are now becoming increasingly important in the characterization of these nano-scale interfaces [21–24]. X-ray reflectometry (XRR) is ideally suited to the study of the internal properties of layered film structures on surfaces, yielding data on sub-surface structure and material properties. The use of neutron reflectometry (NR) in combination with XPS and XRR allows the full chemical composition of plasma-deposited films to be determined.
In this work, we have performed a systematic study of the impact of plasma deposition conditions upon resulting film chemistry, structure and water solvation, which has been correlated with the films' efficacy against protein fouling. The combination of analysis techniques has provided a powerful toolbox for further examination of the specific chemical composition of these plasma polymer films, including hydrogen. Such detailed structure–property correlations may enable a more sophisticated approach to the design of protein-resistant plasma polymer films. Specifically, the surface chemistry of the plasma polymer coatings was characterized by XPS with surface morphology being characterized by atomic force microscopy (AFM). A combination of XPS data with XRR and NR measurements on the plasma polymer film versus air enabled the stoichiometric composition and mass densities of the films to be obtained. Examination of the same films in an aqueous environment using NR and quartz crystal microbalance with dissipation monitoring (QCM-D) highlighted the degree to which the films absorb water. These measurements also showed a substantial exchange of hydrogen atoms between the film and solution, indicating that a large number of hydrogen atoms within the film are labile. Protein adsorption and water uptake studies were performed using QCM-D measurements, and a correlation between the stoichiometric composition of the films (as calculated from reflectivity measurements) and the level of protein adsorption and water uptake was determined. These studies demonstrate the utility of using XRR and NR analysis techniques in furthering the understanding of the chemistry of these films and their interaction with water and proteins.
2. Materials and methods
Ultra-flat single crystal, silicon wafers (〈111〉, 10 cm diameter, 1 cm thick, Silrec Corporation, San Jose, and 〈100〉, 1 cm2 × 0.5 mm thick, from M.M.R.C Pty Ltd, Melbourne, Australia) were used as substrates for the deposition of plasma polymer thin films. Smaller wafers (1 cm2 × 0.05 cm thick) were used as substrates for AFM and XPS characterization. Plasma deposition on the large and small wafers was performed simultaneously. Prior to plasma deposition, the wafers were rinsed with ethanol and cleaned in aqua regia (3 : 1 HCl : HNO3) for an hour to remove any organic and inorganic contaminants. The wafers were further treated with piranha solution (20% H2SO4 in concentrated H2O2) for 3 h to remove residual organic contaminants. This is a standard cleaning protocol that we have developed for the preparation of these surfaces for NR analysis . The wafers were thoroughly rinsed with Milli-Q water and blown dry with nitrogen gas after each step. This procedure did not introduce any measurable roughening of the surface as assessed by AFM (data not shown). The cleaning protocol produces a hydrophilic surface.
2.2. Plasma polymer deposition
Plasma polymerization of di(ethylene glycol) dimethyl ether (99%, BDH Chemicals Ltd) was carried out in a custom-built reactor described elsewhere . Briefly, a cylindrical reactor chamber is used with a height of 35 cm and a diameter of 17 cm. Within this chamber sit two circular electrodes (10.3 cm diameter), spaced 15 cm apart. Samples were placed on the lower grounded electrode and a continuous radiofrequency pulse was generated at the upper electrode. The monomer vapours were supplied to the reactor chamber from the liquid monomer contained in a round-bottom flask via a stainless steel line and a manual valve for fine control of the flow. The monomer flask containing the di(ethylene glycol) dimethyl ether was kept in water at 37°C during the experiments. The monomer liquid was degassed before plasma deposition.
The plasma deposition was performed using a frequency of 125 kHz, load powers of 10, 20 and 50 W and an initial monomer pressure of 20 Pa for a treatment time of 35, 20 and 10 s, respectively, in order to produce films of appropriate thicknesses for reflectivity measurements. The final monomer pressures for the 10, 20 and 50 W plasma depositions were 33, 42 and 60 Pa, respectively. After deposition, the reactor was immediately pumped down to base pressure before venting. The samples were stored in clean tissue culture-grade Petri dishes under ambient conditions until further analysis.
2.3. Atomic force microscopy
An Asylum Research MFP-3D atomic force microscope (Santa Barbara, CA, USA) was used to measure surface topography and roughness in tapping mode with ultrasharp silicon nitride tips (NSC15 non-contact silicon cantilevers, MikroMasch, Spain). The tips used in this study had a typical force constant of 40 N m−1 and a resonant frequency of 320 kHz. Typical scan settings involved the use of an applied piezo deflection voltage of 0.8 V at a scan rate of 0.3 Hz.
2.4. X-ray photoelectron spectroscopy
To investigate the chemical composition of the plasma polymer coatings, XPS was employed. XPS analysis was performed using an AXIS HSi spectrometer (Kratos Analytical Ltd), equipped with a monochromated Al-Kα source at a power of 144 W (12 mA, 12 kV). Charging of the samples during irradiation was compensated by the internal flood gun. The pressure in the main vacuum chamber during analysis was typically 5 × 10−6 Pa. Spectra were recorded with the photoelectron detection normal to the sample surface. All elements present were identified from survey spectra (acquired at a pass energy of 320 eV). High-resolution spectra were recorded from individual peaks (C 1s, O 1s) at 40 eV pass energy (yielding a typical peak width for polymers of 1.0–1.1 eV). The atomic concentrations of the detected elements were calculated using integral peak intensities and the sensitivity factors supplied by the manufacturer. High-resolution C 1s spectra were quantified using a minimization algorithm in order to calculate optimized curve fits and thus determine the contributions from specific functional groups. Five peak components (mixed Gaussian/Lorentzian model functions) were used. Component C1 at the lowest binding energy (BE) was assumed to represent aliphatic hydrocarbons (‘neutral’ carbon) and the corresponding BE set accordingly to 285.0 eV. A second component at a slightly higher BE was included to account for all C 1s photoelectrons that underwent a secondary BE shift. Component C3 at 286.3–286.6 eV represents C–O-based groups (ethers and alcohols), C4 at 287.9–288.2 eV accounts for all C=O and O–C–O-based groups (e.g. carbonyls and amides) and C5 at 288.9–289.3 eV represents O–C=O-based groups (e.g. acids or esters). It is important to note that XPS does not detect light elements such as hydrogen and helium.
2.5. Contact angle measurements
Static water contact angles were determined using a Dataphysics OCA20 goniometer, at 25°C with a drop volume of 1 µl. A photo of the drop was digitized and the profile fitted to the equation of Young and Laplace . Once the drop profile had been determined, the contact angle was calculated from the intersection of the theoretical profile with the baseline. The contact angles presented are the mean of three separate measurements on different regions of the same plasma polymer sample.
2.6. Quartz crystal microbalance measurements
Real-time monitoring of protein adsorption was performed at 25°C using a QCM-D (Q-Sense, Gothenburg, Sweden). Shifts of the oscillating frequency (Δf) were detected and plotted in real time using the resonance frequency at 5 MHz and the third and fifth harmonic. The di(ethylene glycol) dimethyl ether-based plasma polymer films were deposited on gold 5 MHz quartz crystal chips. The chips were cleaned prior to film deposition by immersion in a piranha solution (sulphuric acid, hydrogen peroxide and Milli-Q water (1 : 1 : 5 volume ratio)) and heated to 70°C for 5 min. The crystals were then thoroughly rinsed with Milli-Q water before being dried with a high-pressure stream of purified nitrogen. The plasma polymer-coated chips were first hydrated in the QCM-D over a 1000 min period in phosphate-buffered saline (PBS; flow rate of 10 µl min−1) in order to obtain a stable baseline prior to introduction of the protein. Bovine serum albumin (BSA; Sigma Aldrich) solution in PBS (1 mg ml−1, pH = 7.45) was then flowed through the measuring chamber in contact with the plasma polymer-coated crystal for 60 min, before re-flowing PBS over the coated crystals to wash away any loosely adsorbed protein. The frequency shift (Δf) of the quartz crystal was converted into mass change (Δm) on the electrode surface, calculated using the Sauerbrey equation (equation (2.1), mass sensitivity: 5 ng cm−2) where C = 17.7 ng Hz−1 cm−2 for a 5 MHz quartz crystal and the overtone number, n, is equal to 1, 3, 5, 7. This equation enables an approximate calculation of the amount of protein adsorbed after rinsing, 2.1
2.7. Reflectometry measurements
XRR data were collected using a Panalytical X'Pert Pro instrument (λ = 1.5406 Å). The specular X-ray reflectivity, R (the ratio between the reflected and the incident intensity), was measured over the Q range 0.01 Å−1 < Q < 0.4 Å−1, where Q = 4πsinθ/λ is the momentum transfer and θ is the angle of incidence/reflection. NR data were collected on the NIST NG7 vertical scattering plane reflectometer over the Q range 0.007 Å−1 < Q < 0.15 Å−1. The data were analysed using least squares (differential evolution) in the Motofit program, weighting data on a logarithmic scale and using the instrumental resolution functions. An initial attempt to model the film with a single layer of uniform scattering length density (SLD) indicated that the films were not of a homogeneous composition/density perpendicular to the interface (SLD being dependent on mass density and chemical composition). Therefore, each film was modelled as a composite of up to 12 layers, with each layer having the same thickness, but differing in SLD. For NR datasets, an additional native oxide layer was used (although this is sometimes hard to resolve). Normally, the use of so many layers creates a jagged SLD profile owing to over-parametrization of the system. However, under circumstances in which the SLD gradient is roughly known, it is possible to smooth such a profile by taking the arithmetic mean of the parameters obtained by fitting the data repeatedly. By using the SLD values from the XRR and NR measurements in tandem, it is possible to determine both the mass density and the hydrogen content of each of the sublayers, if one assumes that the atomic compositions from the XPS measurements are constant through the film .
The 20 and 50 W films were also measured against aqueous solution using NR, with the neutron beam reflecting from the film/water surface. Three different solvents (‘contrasts’) were used for the 20 and 50 W films, H2O, D2O and a mix of H2O/D2O (with an SLD of 3.3 × 10−6 Å−2) Changing the deuteration level of the solution changes the refractive index of the water and therefore the scattering contrast of the system. An analysis identical to that carried out on the dry films was performed. By using the average SLD of the dry film, and the SLD of the films in water, the amount of absorbed water can be quantified, as well as the exchange of H atoms from within the film with those of the solvent. For such an analysis, two water contrasts are required.
3. Results and discussion
3.1. Characterization of the films using atomic force microscopy
The surface topography and roughness of the deposited plasma polymer films were determined by AFM in air. Figure 1 shows tapping mode height and phase contrast images of the films in air. The images illustrate the flat, smooth and defect-free nature of these films that make them ideal for studies using reflectivity measurements. Analysis of the images reveals that all of the films were very smooth (root mean square (RMS) values between 0.45 and 0.57 nm). The size of the topographical height variations in the X and Y dimensions are of the order of 20–50 nm. The film deposited at 10 W displayed minimal phase contrast with the 20 W being intermediate between the high- and low-power films. The static water contact angles of the three films produced at 10, 20 and 50 W were measured immediately after deposition of each film, and were found to be very similar at 65° ± 2°, 60° ± 4° and 66° ± 4°, respectively. The contact angles observed for these films are similar to the values reported in a recent study of di(ethylene glycol) dimethyl ether plasma polymer films deposited on silicon at plasma load powers of 10 and 20 W by Cheng & Komvopoulos .
3.2. Film stability, water uptake and protein-resistant properties of the plasma polymer films characterized by quartz crystal microbalance with dissipation monitoring
The protein-resistant nature of the three films was evaluated against BSA using QCM-D. Initially, the films were pre-equilibrated in a flow of PBS for 1000 min to allow them to fully hydrate and remove any loosely bound, low molecular weight material before incubation with protein (1 mg ml−1 at pH 7.4). Figure 2a shows the QCM-D frequency responses of the films over the 1000 min hydration period. A substantial loss of low molecular weight material is seen in the 10 W film as shown by the increase in frequency response over time. It is not uncommon for films produced at lower powers or under pulsed conditions to result in the formation of such low molecular weight material . The 10 W film rapidly loses material over the first few hours of PBS flow and then appears to stabilize. Conversely, a decrease in frequency response over time is observed for the 50 W film, indicating a substantial amount of mechanically coupled mass (water and buffer salts). Interestingly, from QCM-D analysis, the 20 W film appears to be very stable during the flow of PBS solution and no significant increase or decrease is observed in its frequency response over time. This finding will be discussed further when NR data in water of the 20 W plasma polymer film are reported. Analysis of the dissipation response (figure 2b) of the 10, 20 and 50 W films over the 1000 min hydration period confirms the higher uptake of water seen in the 50 W film, with increasing dissipation (ΔD = 8 × 10−6) compared with no dissipation change seen for the 20 W film. Furthermore, a decrease in the dissipation (ΔD = −30 × 10−6) of the 10 W film not only represents a more rigid film than the 20 and 50 W films but also reflects a lower amount of absorbed and coupled water compared with the 20 and 50 W films.
The frequency response, indicative of the mass change of the three plasma polymer films after flowing a solution of BSA, is shown in figure 2c. It can be seen that the level of protein adsorption clearly increases with plasma deposition power. If one makes an approximation on the amount of protein adsorbing in the films using the Sauerbrey equation (equation 2.1), we calculate the amounts of protein fouling to be 15, 60 and 80 ng cm−2 on the 10, 20 and 50 W plasma polymer films, respectively. We acknowledge the limitations of using this equation in accurately quantifying protein adsorption within these plasma polymer films. Bretagnol et al . has reported a similar study of protein adsorption on diglyme plasma-polymerized surfaces. The level of BSA adsorption on the DGpp films deposited at 1, 5 and 15 W was measured using QCM-D, and using the Sauerbrey equation they reported BSA adsorption of approximately 20, 25 and 110 ng cm−2, respectively, compared with approximately 600 ng cm−2 measured on the control SiO2-coated quartz crystal .
In addition to QCM-D analysis, protein adsorption experiments were performed on the plasma polymer surfaces at the same concentration of BSA (1 mg ml−1) for 1 h followed by repeated rinsing in PBS and then Milli-Q water. An emerging nitrogen (N) signal, measured by XPS, was used to monitor protein adsorption on the surfaces. The atomic% nitrogen values in the films after BSA incubation were 0.6 ± 0.2%, 2.7 ± 0.2% and 4.3 ± 0.4% for the 10, 20 and 50 W films, respectively (data not shown). It is clear that significantly less protein absorption was observed on the 10 W film than on the higher power films. In another study reported by Bretagnol et al. , BSA adsorption on a diglyme plasma-polymerized surface deposited under pulsed plasma discharge conditions yielded a value (0.5%) that was similar to that reported for the 10 W film in this study (0.6%). This was compared with BSA adsorption measured on a silicon wafer control, which yielded 12 per cent N content after incubation in BSA (100 µg ml−1 for 1 h). A comparison of the C–O (ether related) component, as measured by XPS, showed that the diglyme surface in Bretagnol's study was higher, at 73 per cent, than the 10 W film reported in this study, which was 62 per cent . However, Salim et al . have reported substantially lower protein adsorption on tetraglyme plasma-coated microfluidic channels, using pulsed plasma discharge conditions with N 1 s levels of 0–0.1% after incubation in a 50 µg ml−1 solution of human fibrinogen. The low protein adsorption in this study compared with those reported in the current study could be attributed to a higher C–O component (84%), lower protein concentrations and differences in plasma-processing parameters such as the pulsed plasma conditions and the use of a tetraglyme monomer . Previous studies have shown that tri- and tetraglyme plasma polymer films are extremely effective in reducing surface protein adsorption [15,20].
3.3. Surface chemistry
The results of XPS elemental analysis of freshly deposited films of di(ethylene glycol) dimethyl ether plasma polymers at load powers of 10, 20 and 50 W are summarized in table 1. The corresponding quantitative results and atomic ratios relative to total carbon are also compiled in table 1. Elemental analysis by XPS reveals films comprising both carbon and oxygen with atomic concentrations varying from 69 to 76 per cent and from 31 to 24 per cent, respectively, across the load power range consistent with previous studies of these di(ethylene glycol) dimethyl ether plasma polymers . The chemical composition of the 20 and 50 W films differs substantially from the monomer (C2H4.7O1), for which the oxygen content is higher (33% O, 67% C). It is clear that higher power glow discharges result in the deposition of plasma polymer films that contain less oxygen and a greater degree of hydrocarbon-containing species. The absence of silicon in the recorded spectra suggests a plasma polymer film thickness in excess of the XPS analysis depth of 10 nm. Fitting of the high-resolution carbon (C 1s) spectra (figure 3) reveals films that are rich in carbon–oxygen moieties, such as ether, alcohol, aldehyde, ketone, ester and acid species. The presence of a higher C–O component suggests the incorporation of a high level of ether functionality in the low-power films consistent with recent work investigating the gas phase discharge of di(ethylene glycol) vinyl ether and di(ethylene glycol) dimethyl ether plasmas [27,31–33]. The ionization/dissociation rate of the monomer ether units increases dramatically with plasma power, resulting in films that contain significantly less ether units. The 10 W film retained the highest C–O component compared with the 20 and 50 W films. This indicates that depositions performed at 10 W load power, are more effective in retaining the ether functionality of the starting monomer, when compared with the other deposition conditions investigated in this work. Films deposited at 20 and 50 W load powers had a higher introduction of neutral hydrocarbon (C–C/C–H) species and a decreased retention of C–O ether and alcohol functionalities.
3.4. Characterization of the films using X-ray and neutron reflectometry in air
The air–solid reflectivity measurements from the plasma polymer films are shown in figure 4, while the structural parameters of each film are given in table 2. The large number of Kiessig fringes in the X-ray reflectivity data for 10 and 20 W which persist to high Qz (=0.4 Å−1) indicates that the films are smooth and provides confidence for the precise determination of film thickness. However, these fringes fade quickly for the 50 W film, indicating that there is a steep gradient in SLD through the film. This is consistent with the observation of a rapid increase in pressure of 40 Pa during the plasma deposition of di(ethylene glycol) dimethyl ether at a load power of 50 W over 10 s. As the pressure is rapidly increasing during plasma polymer deposition of the 50 W film (4 Pa s−1), it would be expected that a significant gradient in film composition with thickness would occur in this plasma polymer thin film, which was deposited at a rate of approximately 2 nm s−1. For comparison, the 10 W load power film only resulted in a 13 Pa increase in pressure over 35 s correlating to a pressure change of 0.4 Pa s−1 and a deposition rate of approximately 1 nm s−1.
The coherent neutron scattering length (b) of H (−3.739 fm) is significantly different (both in sign and magnitude) from that of the other atoms (C, 6.646 fm; O, 5.803 fm) found in these plasma polymers. As the hydrogen content of these plasma polymer films is increased, the neutron SLD decreases much more quickly than the X-ray SLD (at constant molecular volume). However, both SLDs are proportional to mass density. By simultaneously fitting the composition and mass density to the average X-ray and neutron SLD values determined for these plasma polymer films along with the XPS results (table 1; XPS is not sensitive to H content), it is possible to calculate the average density of the film and the full atomic composition for the films (table 2). There is an assumption involved in these calculations, namely that the surface composition (atomic ratios) determined by XPS is similar to those in the bulk. The first layer (closest to the air) was ignored in the calculation of the average film properties. This is because the film roughness can be coupled to the mass density/atomic composition. Thus, if the fit underestimates the roughness of the film, then the fitting process can compensate by having a low mass density, or high H content, for the surface layer. It is clear that, for all plasma polymers produced, there is a loss of oxygen and most significantly of hydrogen during the radio frequency glow discharge plasma polymerization of the di(ethylene glycol) dimethyl ether monomer. This loss of hydrogen during the deposition of plasma polymers has been observed by us previously for other reactive functional monomer species [23,24]. The differences between the resulting plasma polymer film chemistry and the starting monomer composition are most significant for the higher power films, for which there is the most hydrogen depletion.
Interestingly, the density of poly(ethylene glycol) polymers varies from around 1.1 to 1.2 g cm−3 and increases with molecular weight. The density of the plasma polymer film produced at 10 W that was found to be the lowest protein-fouling film is within the reported density range of this class of polymer at 1.19 g cm−3. The atomic composition of the 10 W film (C2H3.3O0.9) was closest to that of the starting monomer (C2H4.7O1) and similar to that of PEG polymers (C2H4O1). This highlights the possible similarities between both the chemistry and structure of the lowest power plasma polymer thin film and the chemistry and structure of PEG graft polymers. The outermost layer of the 10 W film also has the greatest H/C ratio near the air interface of all three plasma polymer films analysed in this study, which decreased systematically with power at the air interface. We therefore infer that the 10 W film is the most PEG-like in terms of film chemistry of the three films studied in this work, in particular at the surface (air interface) where interactions with proteins (when the films are placed in solution) are most relevant from both XPS and reflectivity measurements in air.
It can be seen from figure 5 that the 50 W film has the lowest H/C ratio of all the films, which is to be expected considering that a greater degree of monomer fragmentation occurs during the glow discharge polymerization of di(ethylene glycol) dimethyl ether at 50 W. This is evident from the increase in pressure and deposition rate during plasma polymerization at this load power compared with the lower power films. Interestingly, the 50 W film has the lowest average mass density of all the films produced. Strikingly, the mass density of the 50 W film at the air interface (the first 5 nm) is dramatically lower than that of its bulk and of the bulk of both the 10 and 20 W films (figure 5). This may, in part, be due to the rougher surface of the 50 W film. It is also possible that there may be a substantial amount of polymerization occurring in the higher powered plasma glow, within the plasma reactor system. If this was to occur, it is likely that the longer chain species could deposit on the silicon substrate during plasma polymerization and this may result in a less dense film being produced. In the light of this finding, and taking into account the film chemistry of the 50 W film, it is not surprising that the 50 W film adsorbs a greater amount of protein than the lower power films.
The reflectivity data for the 20 and 50 W films against various H2O/D2O mixes are shown in figure 6. From inspection of the 20 W film in each of the D2O/H2O mixtures, it is obvious that the films absorb a significant amount of water, as the SLD of the film changes substantially. By using the average SLD of the film in H2O and the average SLD scattering length of the film against air, one can calculate the amount of absorbed water in the film. For the 20 W film, this corresponds to 22% v/v water ingress. For the 50 W film, the amount of absorbed water is much higher at 40% v/v. When placed in water, both films swelled by a significant amount—a 9 per cent increase in film thickness for the 20 W film was observed, and a 10 per cent increase in film thickness for the 50 W film was measured. It is important to note here that QCM-D measurements (figure 2a) confirmed that the 50 W film absorbed a substantial amount of water, but this was not indicated from QCM-D measurements in the 20 W film. This has important implications in the use of QCM-D measurements in probing the hydration of plasma polymer films. We propose that in the 20 W film, as no increases in mass were detected after incubation in PBS, the amount of low molecular weight material being removed in this film over time was balanced by water ingress into the film. The NR measurements in water in combination with QCM-D measurements provide strong evidence of this mechanism. It is not uncommon for films produced at lower powers or under pulsed conditions to result in the formation of such low molecular weight material. The implications of this low molecular weight material, in terms of the lower power films, protein-resistant nature, may be significant. The fact that the 50 W film adsorbs substantially more water than the film deposited at 20 W is interesting. Both films have very similar water contact angles and, therefore, surface hydrophilicity. We propose that the discrepancy in water absorption is due to the lower average mass density of the 50 W film. This film, therefore, allows for greater water adsorption and diffusion throughout its bulk when compared with the 20 W film with the greater mass density. The level of water absorption in these films is far higher than we have observed in previous NR and XRR studies of amine-containing plasma-polymerized allylamine films . In comparing our previous study of an allylamine plasma polymer film that had a water contact angle of 55° and a fitted mass density of 1.46 g cm−3, it absorbed only 3% v/v water compared with 40% v/v for the 50 W di(ethylene glycol) dimethyl ether plasma polymer film, which has a similar water contact angle of 66° and a significantly lower mass density of 0.99 g cm−3. This previous work would seem to support our hypothesis that the mass density of these di(ethylene glycol) dimethyl ether plasma polymer films plays a significant role in their water uptake. It appears that water contact angle measurements are not a good determinant of the solvation properties of these thin films. One cannot rule out the important influence of the film chemistry in addition to their mass densities in the extent of hydration of these plasma polymer films. It is well known that the ether units in PEG-based polymers are extremely good at hydrogen bonding and are extremely well hydrated under most conditions. By this rationale, however, one would expect the plasma polymer deposited at 20 W to absorb more water than the 50 W film as it contains a higher concentration of ether groups, which is not the case.
Since we measured the 20 W film against two other D2O/H2O mixtures, it is also possible to calculate the proportion of protons that exchange with solution. For the 20 W film, approximately 15 per cent of the protons exchange with solution. In general, labile hydrogen atoms are more commonly found on carboxyl and hydroxyl species that we have previously shown to be present in these plasma polymer films . Hydrogen atoms bonded to carbon are generally much less labile. From analysis of the XPS C 1s high-resolution spectra curve fits (table 1), it can be seen that a minimum of 13 per cent of the oxygen groups in the 20 W film are not related to ether species. It is reasonable to assume that a small component of the C–O curve fitted species is hydroxyl species, as shown in our previous work. Therefore, the fact that we are seeing 15 per cent of the protons in the 20 W film can be rationalized by taking into consideration its significant degree of solvent penetration and the observation of residual ‘non-ether’ chemical species within the film. Unfortunately, this analysis cannot be repeated for the 50 W film, as the D2O and D2O/H2O solvents possessed similar SLDs to the swollen film.
The proposed mechanisms of non-specific protein adsorption include electrostatic, van der Waals, hydrophobic and hydrogen-bonding interactions . The 10, 20 and 50 W films had very similar static water contact angles of approximately 60°. It appears that, with regards to the di(ethylene glycol) dimethyl ether plasma polymer films, surface tension effects are not a significant discriminating factor in their relative water solvation and protein-fouling characteristics, as has been previously reported [15,31]. In a study on plasma polymer films deposited from ethylene glycol-containing monomers, Johnston et al.  have shown that molecular surface structure primarily affects the ability of PEG-like plasma polymer films to resist protein adsorption. Little correlation was observed between the contact angles measured and protein adsorption, suggesting that surface tension and interfacial tension effects are not the primary factors that influence protein adsorption and water solvation in these types of plasma polymer thin films.
Based on the XPS data presented in table 1, there are some clear differences in the composition of the di(ethylene glycol) dimethyl ether plasma polymer coatings prepared under different deposition conditions. The 10 and 20 W films are high in ether (C–O) carbon species. However, the composition of the 50 W film was very different from that of these lower power films and probably relates to the degree of monomer fragmentation. This is supported by the observations of pressure rise and deposition rates during the plasma polymerization of di(ethylene glycol) dimethyl ether at different powers. Prior to incubation in the protein solutions, one plasma polymer coating contained a high C–C/C–H content of over 50 per cent (50 W) and two coatings contained a high C–O content (10 and 20 W) of over 50 per cent. The relative amount of BSA adsorption was significantly reduced on the 10 W film (figure 2), which was the most PEG-like. A similar result was found by Bretagnol et al.  in a study of di(ethylene glycol) dimethyl ether plasma polymer thin films produced at load powers of 1 and 5 W.
It is becoming clear that chain density and, thus, conformation are critical factors in the successful low-protein-fouling properties of PEG and PEG-like films [19,35–37]. In a study by Fick et al. , PEG self-assembled monolayers were reported to be highly protein resistant when they contain a solvated densely grafted brush of only 40 Å thick, which correlates to a chain length of only 11 ethylene glycol units. Our fitted reflectivity measurements point to the presence of such thin PEG-like plasma polymer films when they are produced at lower powers, and QCM-D measurements in correlation with NR measurements in water have shown that the films lose mass in water owing to the presence of low molecular weight material. It is reasonable to assume that the 10 and 20 W films are cross-linked to some extent, with a high degree of residual ether-containing chains at the surface and some low molecular weight material. Although we could not unambiguously confirm it experimentally, we expect the 10 W film to contain a greater amount of low molecular weight material than the 20 W film. We hypothesize that the reasons why these particular films are lower fouling than the 50 W film are the higher density of residual ether-containing functional groups in these films, along with the film structure appearing similar to that of PEG graft surfaces and perhaps some low molecular weight material. It is clear from XPS analysis that the 50 W film has the lowest concentration of surface ether groups and the highest hydrocarbon component, which was also observed in the neutron and XRR compositional fits. Therefore, the reason for its greater propensity to adsorb proteins must be at least in part due to this greater hydrocarbon component and oxygenated species when compared with the lower power films.
Thin films of di(ethylene glycol) dimethyl ether have been generated by plasma polymerization and studied using XPS, QCM-D, X-ray and NR. Using combinations of these techniques, we were able to accurately determine the film thickness and RMS roughness and to decouple the films' composition from their mass densities. All films result in a substantial loss of hydrogen when compared with the starting monomer, with H/C values at the air interface being higher in films deposited at lower load power. Similarly, XPS and reflectivity measurements showed that the chemical composition of the 10 W film was most similar to that of PEG-grafted surfaces and to the starting monomer, particularly at the surface (air interface) where the interactions with proteins are dictated (when the films are placed in solution). The films were found to absorb a significant amount of water (approx. 22–40% v/v) and their degree of solvation appears to be dependent on both the film chemistry and mass density profile. Films deposited at higher load powers resulted in lower mass densities and hence higher amounts of absorbed water. At the air interface, the mass densities of the 10 and 20 W films were most similar to PEG polymers, while the 50 W film had a very low mass density. A combination of the lower mass density and less PEG-like surface chemistry of the 50 W plasma polymer film was correlated with the higher amount of protein adsorbed, compared with the 10 and 20 W films. Surface hydrophilicity was shown to be a poor determinant of solvation and protein resistance in these thin films. A substantial proportion of the hydrogen atoms within the di(ethylene glycol) dimethyl ether plasma polymer film deposited at 20 W were ‘reactive’ and exchangeable with solution, explaining their high degree of absorbed water. QCM-D measurements were found to be inaccurate in detecting associated water in the lower power films that contained loosely bound, low molecular weight material. The lower powered plasma polymer film that contained a high ether content displayed significantly reduced adsorption of BSA from the concentrated protein solution. Plasma polymer films with low residual ether content that contained low molecular weight material adsorbed much greater amounts of protein from solution. Reflectivity measurements show that the film with the greatest chemical similarity to that of poly(ethylene glycol)-grafted surfaces was the least fouling to proteins. In addition, the low-fouling 10 W film was also found to have the highest H/C content at the air interface. Therefore, the higher protein resistance of the di(ethylene glycol) dimethyl ether plasma polymer film deposited at a plasma load power of 10 W was attributed to its greater chemical and structural similarity to that of poly(ethylene glycol) graft surfaces. Interestingly, the degree of protein adsorption did not positively correlate with films that absorbed higher amounts of water, as is often suggested by the water barrier theory. This work shows the utility of combining NR and XRR measurements of plasma polymer thin films in parallel with techniques such as XPS and QCM-D.
We thank the CSIRO OCE scheme for funding the PhD scholarship for D.J.M. and the Access to Major Research Facilities Project (AMRFP) for funding. We thank the National Institute of Science and Technology, USA, for providing the grant time to use the neutron reflectometer (NG7) as well as Hanna Wacklin (ESS) and Bulent Akgun (NIST) for assisitance in attaining the NR data.
- Received August 1, 2011.
- Accepted September 7, 2011.
- This journal is © 2011 The Royal Society