1. Introduction
Organic thin film materials have many potential and implemented applications, from biocompatible and antifouling coatings in medical devices to protective coatings, waveguides and component materials in optoelectronic devices, such as flexible organic electroluminescent devices (OLED), organic photovoltaics (OPV) and organic thin film transistors (OTFT) [
1,
2,
3,
4,
5,
6]. The motivation for using organic materials stems from their chemical versatility, low cost, low temperature fabrication and ability for reel-to-reel printing, as well as mechanical flexibility [
1]. The attractive synthetic and processing flexibility of organic materials allows for fine tuning of their properties to achieve a desired combination of optoelectronic, physicochemical, mechanical and biological characteristics, and controlling the manner in which these materials interact with relevant liquids [
7]. The latter is an important aspect for optoelectronics and biomedical applications alike. Solution processing of electronic circuits demands dielectric interlayer stability in organic solvents and strong interfacial adhesion, while the susceptibility of the coating to the uptake of moisture is critical for OPV encapsulation applications. Surface wetting by physiological fluids is a key determinant of biocompatibility and degradation under
in vitro and
in vivo conditions [
1,
8]. The surface-solvent kinetics and ensuing degradation of organic materials under aqueous conditions is an important consideration for biodegradable and compostable electronics, where it would affect both operational performance and the physical transience of the device [
9].
One of the key predictors of surface-solvent interactions, the wetting behaviour at polymer surfaces and interfaces, is dependent on the nature of the variable length of polymer chains, density fluctuations and the relatively slow motion of the long chain molecules [
10]. Contact angle (CA) analysis is typically used to obtain primary data from which the degree of wettability for a specific solid-liquid combination can be inferred [
7,
11], with lower CAs indicating better wetting compared to larger CAs. This analysis also provides indirect information about the structure of the film matrix, where observed changes to θ at the liquid/solid interface are attributed to specific solid-liquid interaction mechanisms, such as absorption, spreading and swelling [
12]. Fundamentally, wettability is affected by the chemical composition, topography, rigidity and homogeneity of the surfaces. The chemical composition of the surface determines its surface energy: whereas surfaces rich in non-polar groups (e.g., –CH
x, with
x = 1–3) have low surface energies and, thus, are hydrophobic, surfaces with a high density of polar groups (e.g., –OH or –C=O) exhibit high surface energies and are hydrophilic [
13]. With regard to surface topography, an increase in the surface roughness enhances the surface hydrophilicity of hydrophilic materials and hydrophobicity in the case of hydrophobic surfaces [
14]. The micro- and nano-scale variations in the slope on the surface are believed to create physical barriers that directly affect the motion of the contact line, thus affecting CAs observed at the macro scale. In a similar way, distinct domains of chemically heterogeneous surfaces, e.g., those with higher hydrophobicity, are thought to interfere with the motion of the contact angle by hindering the advancement or contraction of the water front and, thus, increasing or decreasing the observed CA. Several models propose a relationship between the measured CAs of a given non-ideal surface and its flat, homogenous counterpart of the same composition, among them Wenzel and Cassie–Baxter models [
15,
16]. In the Wenzel model, the liquid is assumed to be in contact with all of the parts of the irregular surface and is typically applied to chemically homogenous surfaces, whereas the Cassie-Baxter model places the drop on the surface protrusions without wetting the entire surface and is believed to be more appropriate for chemically heterogeneous surfaces.
Recently, growing interest in environmentally-friendly technologies has led to the exploration of a number of alternative organic source materials, e.g., unprocessed raw agricultural, food and waste substances [
17], and energy-efficient green fabrication methodologies, e.g., plasma-assisted nanofabrication [
18,
19]. The key driver of this research therefore rests in the need to support robust and sustainable economic and societal development, which, in practical terms, means the development of materials and technologies that are cheaper, more efficient and can address the objectives of modern societies in an environmentally-sustainable fashion. Most of the existing chemical synthesis processes used for the fabrication of electronics device components are energy-inefficient and require multi-step processing and the use of hazardous auxiliary substances, such as organic solvents and catalysts [
20]. These synthesis routes often rely on expensive and/or toxic, high purity, non-renewable materials [
1]. In addition to using hazardous materials, modern electronic devices also use valuable and scarce materials, the availability of some of which (e.g., gallium, indium,
etc.) is reducing at a high rate [
1]. At the same time, only a limited portion of high-tech waste is recycled, with the bulk of the waste being deposited into landfills, where it slowly degrades, leaching out potentially harmful by-products. By using minimally processed, renewable natural resources, such as non-petrochemical oils, e.g., essential oils, and highly reactive non-equilibrium dry plasma-based chemistry, it is possible to lower the environmental footprint and the economic costs of organic electronics and other such technologies throughout their lifecycle, from their fabrication to their use and disposal [
21].
Essential oils are volatile aromatic compounds that are widely used in pharmacological, perfumery and culinary preparations for their aromatic and medicinal properties. Their availability in commercial quantities, relative low-cost, renewable nature and minimal toxicity makes them a suitable precursor for “green” functional materials [
22]. As a volatile material, essential oil is well suited to chemical vapour deposition (CVD), as no carrier gas is required to deliver the monomer into the polymerization chamber. Using plasma as a catalyst, the oils can be converted into functional polymer thin films in a one-step process at room temperature, without the need for pre- or post-processing, e.g., annealing, or catalysts. Unlike most plasma deposited films, plasma polymers of essential oils are optically transparent and smooth. Over the last few years, thin films from lavender and tea tree essential oils have been developed and identified as promising candidates for applications in electronics, as dielectric and encapsulation layers [
23,
24,
25,
26,
27]. Plasma polymers of terpinen-4-ol also display a valuable and rare electron blocking hole-transporting property that is very attractive for OLEDs [
27]. Versatile biological activity, including antibacterial activity against drug-resistant strains of
Staphylococcus,
Streptococcus and
Candida species, also makes essential oils an attractive candidate for the fabrication of antibacterial and biocompatible implantable systems, including implantable electronics [
28].
This paper reports the fabrication of polymer thin films from 1-isopropyl-4-methyl-1,4-cyclohexadiene, also known as γ-Terpinene, using RF plasma polymerization. γ-Terpinene is an isomeric hydrocarbon distilled directly from
Melaleuca alternifolia essential oil and used extensively in cosmetics and cleaning products for its aromatic and medicinal properties. While γ-Terpinene has been determined as a component in plants using many quantitative analysis methods [
29,
30], it had not been studied extensively as the main topic compound [
31]. However, recent investigations of γ-Terpinene revealed that the optical properties of the fabricated plasma-polymerized γ-Terpinene (pp–GT) thin films are promising, with transparency to the optical wavelengths and refractive index of 1.57–1.58 (at 500 nm) [
26]. It is reported that these polymers possess an optical band gap (
Eg) of ~3 eV that falls into the insulating
Eg region. Independent of deposition conditions, the surfaces are smooth and defect-free, with uniformly distributed morphological features and average roughness well below 0.30 nm [
26]. The optical and surface characteristics suggest that the pp–GT thin films have the potential to be implemented in optoelectronic and insulating applications. This paper focuses on the wetting, solubility and chemical characteristics of these films and the compatibility of the material with solvents typically utilized in the manufacturing of organic electronic devices, with the intention to use these films in OPVs, especially as encapsulation coatings and insulating layers in flexible electronics.
2. Experimental Section
2.1. Preparation of Thin Films
Thin film samples were deposited on high quality glass microscope slides using a custom-made RF polymerization chamber, 0.75 m in length with an inner diameter of 0.055 m (volume of 0.0035 m3). RF power is delivered to the system via an ENI RF generator at 13.56 MHz through a matching network and capacitively coupled copper electrodes. The copper electrodes were placed 0.11 m from the monomer inlet and 0.1 m apart, with the active electrode closest to the monomer inlet. The electrode configuration used was based on the uniformity of the RF discharge it produced and the corresponding uniformity of resulting thin films for the particular reaction chamber utilized throughout this work.
Prior to deposition, the substrates were thoroughly washed in extran, cleaned ultrasonically in water, rinsed in isopropanol and distilled water, air dried and placed in the reaction chamber. For each deposition, 5 mL of γ-Terpinene (GT) monomer was used. The chamber was evacuated to a pressure of ~100 mTorr, at which stage the monomer inlet was opened to allow the monomer to evaporate. Argon gas was then used to flush the chamber for 1 min at a pressure of 1000 mTorr to remove residual background gas and ensure an oxygen-free surface. The chamber was then evacuated to 100 mTorr, at which stage the monomer inlet was closed, and RF glow (10, 25, 50 and 75 W) was initiated. The monomer-free plasma state was maintained for 2 min to stabilize the pressure, as well as to etch potential residual contaminants from the surface of the substrate. Once the pressure had reached 150 mTorr, the monomer inlet was opened, beginning the deposition. The flow rate was controlled via a vacuum stopcock and was estimated to be 1.57 cm
3/min by employing the procedure outlined by Gengenbach and Griesser [
32].
Following the aforementioned experimental procedure, plasma polymer thin films were fabricated from γ-Terpinene (GT) monomer (
Figure 1) at various RF power levels (10, 25, 50 and 75 W) and an ambient temperature of 20 °C. pp–GT thin-films were examined over the wavelength range 190–1000 nm using a variable angle spectroscopic ellipsometer (model M-2000, J.A. Woollam Co., Inc., Lincoln, NE, USA). Ellipsometric parameters Ψ and Δ were obtained at three different angles of incidence, φ = 55°, 60° and 65°. In addition, the transmission data were also collected. Ψ and Δ were used to derive the optical constants based on a multilayer model consisting of a previously modelled substrate and Cauchy layer built in the J.A. Woollam Inc. analysis software (WVASE32) [
33] via regression analysis. The quality of the fit was measured quantitatively by determining the mean-squared error and through the use of the correlation matrix. Gaussian oscillators were employed within the model to provide an optimal fit of the data, with a lower mean square error and lower average correlation between fitting terms. A more detailed review of the procedure has been reported elsewhere [
34]. Samples for Fourier transform infrared (FTIR) and CA measurements were deposited for 30 min to obtain films of ~700 nm thickness. Depositions were performed for 45 min to obtain films of ~1 μm for atomic force microscope (AFM) measurements.
Figure 1.
Conformers of 1-isopropyl-4-methyl-1,4-cyclohexadiene resulting from the rotation about the C
7–C
1 bond, adapted from [
35].
Figure 1.
Conformers of 1-isopropyl-4-methyl-1,4-cyclohexadiene resulting from the rotation about the C
7–C
1 bond, adapted from [
35].
2.2. Chemical Characterization
FTIR spectroscopy was carried out for the chemical characterization of the γ-Terpinene monomer and the pp–GT thin films using a Perkin Elmer Spectrum 100 FTIR spectrometer. Spectra were obtained in transmission mode in the region of 4000–500 cm−1, where 32 scans were acquired for each sample at a resolution of 1 cm−1. Contributions from CO2 and H2O were eliminated from the spectra by a background subtraction procedure, where the background was pre-recorded under the same atmospheric conditions.
The surface chemistry of the deposited samples was further analysed by XPS on a SPECS SAGE XPS system equipped with a Phoibos 150 hemispherical analyser and an MCD-9 detector. The background pressure was held at 2 × 10
−6 Pa during experiments. For wide scan spectra and high resolution scans of the C 1s peak, an Mg Ka X-ray source was used (h
v = 1253.6 eV), operated at 10 kV and 20 mA (200 W). Measurements were performed with a pass energy of 100 eV, and 0.5 eV energy steps were used for wide scan spectra, while 20 eV pass energy and 0.1 eV energy steps were used for high resolution scans. Spectra were analysed using CasaXPS (Case Software Ltd, Teignmouth, UK). Synthetic peaks were fitted to the C 1s envelops following the methodology of Beamson and Briggs [
36] with the spectra correction for charging effects during analysis using a reference value of 285 eV, the binding energy of the C–C component from neutral hydrocarbon [
37]. The analysis area was circular with a diameter of 5mm, and spectra were acquired at a take-off angle of 90°.The full width at half maximum (FWHM) of the C 1s synthetic peaks remained constant at 1.5 eV.
2.4. Surface Free Energy Analysis
There are several widely applied approaches for the determination of the solid surface free energy and its components from CA measurements [
40]. Most theories of solid surface energy have a basis in Young’s equation that employs the equilibrium CA, where the solid is considered close to ideal. The ideal surface is one that is chemically and morphologically homogenous, and thus, CAH is assumed to be absent or negligible.
For the non-ideal surfaces that are chemically and morphologically inhomogeneous, with a measurable, substantial CAH, the apparent surface free energy γ
SV and other interfacial interaction parameters, adhesive film tension
Π, work of adhesion
WA and work of spreading
WS, can be derived from the CAH approach developed by Chibowski [
41] with only three measurable quantities: the surface tension of the probe liquid γ
LV and its advancing θ
A and receding θ
R CA hysteresis (
CAH = θ
A − θ
R). The solid surface free energy γ
SF can be expressed by the following relation [
41]:
The apparent, total surface free energy of a solid γ
SV (≈ γ
SF) can be expressed as [
41]:
CAH can be related to the work of spreading
WS of liquid on the polymer surface.
WS is a thermodynamic quantity that relates the wettability to the mechanical strength of adhesion. It enables one to characterize the competition between solid-liquid adhesions with different liquids [
40]. W
S can be easily calculated from the work of adhesion
WA and the work of cohesion
WC:
where
WA = γ
LV (1 + cosθ
A) and
WC = 2γ
LV [
40].
For surfaces that are chemically heterogeneous, but very smooth, the experimentally observed advancing CA θ
A might be expected to be a good approximation of Young’s CA θ
Y, whereas the experimental receding angle, θ
R, is expected to have less reproducibility, due to liquid sorption or solid swelling [
39]. With the assumption of θ
A = θ
Y = θ, the surface free energy can be calculated by means of the van Oss’ adaptation of Young’s theory [
42,
43]. The quantitative determination of all of the surface thermodynamic properties of the polymer coatings was performed using the Young-Dupré equation [
43]:
where θ
a is the advancing CA (°), γ
L is the surface tension (SFT) of the liquid in contact with the solid surface (mJ/m
2),
is the apolar component (Lifshitz–van der Waals [LW]) of the SFT of the liquid (mJ/m
2),
is the electron-acceptor parameter of the polar component (acid–base [AB]) of the liquid (mJ/m
2),
is the electron-donor parameter of the polar component (AB) of the liquid (mJ/m
2),
is the apolar component (LW) of the surface energy of the solid (mJ/m
2),
is the electron-acceptor parameter of the polar component (AB) of the solid (mJ/m
2) and
is the electron-donor parameter of the polar component (AB) of the solid (mJ/m
2).
The interface interaction of the polymer and the solvent is determined from their interfacial tension γ
12 using the following equation [
43]:
where Δ
G121 is the free energy change. In the case of two completely miscible substances, where the interfacial tension cannot be measured directly, γ
12 can be calculated using the following equation:
In case of two completely immiscible substances (or any given solid-liquid system), the interfacial tension can be derived directly from the measured contact angle θ:
There are several other methods for obtaining surface tension values; however, the appropriateness of these methods for the probing of particular polymer/solvent combinations remains a subject of debate [
44,
45,
46]. Fowkes [
47] and Neumann’s [
44,
48] approaches were chosen in this study to provide a basis for comparison.
Fowkes method is commonly used for the determination of the surface free energy of polymeric materials [
49]. Two-phase systems are investigated that contain a substance (solid or liquid) in which only the dispersion interactions appear. Considering such systems, Fowkes determined the surface free energy corresponding to the solid-liquid interface using the following equation [
49]:
According to Fowkes [
50,
51], the combination of Equations (8) and (9) yields the formula that enables one to calculate the surface free energy of a solid for which γ
S =
is valid [
49]:
If the measuring liquid is a dispersive one (γ
L =
), Equation (10) simplifies to:
To determine γ
S of any solid, the CA for the solid is measured using the dispersion liquid. Then,
is calculated from Equation (11). Next, the CA (θ
p) is measured using a liquid for which γ
L =
+
. The
can be estimated using [
49]:
Fowkes method is based on the independence and additivity of the dispersion and polar interactions [
49]. On the other hand, Neumann’s approach derives the SFT from a purely thermodynamic point of view and, therefore, neglects the molecular origins of SFT [
7]. However, this is the only theory that allows the calculations to be done by using just one probing liquid. The following equation provides a method for calculating the surface energy of a solid from a single CA value [
44,
48]:
where β is an experimentally-derived constant (β ≈ 0.0001247) [
49].
2.5. Determination of Roughness and Wetting Behaviour of Surfaces
To analyse the effect of surface roughness on the wetting properties of the pp–GT thin films, the polymer was deposited on glass, silicon (Si), silicon dioxide (SiO2) and indium tin oxide (ITO) substrates using RF plasma polymerization at 75 W RF power. The surface morphology and roughness parameters of the films were determined from AFM images acquired on a NT-MDT NTEGRA Prima AFM operating in semi-contact (tapping) mode, using (NSC05, NT-MDT) cantilevers with a spring constant of 11 N/m, a tip radius of curvature of 10 nm, an aspect ratio of 10:1 and a resonance frequency of 150 KHz. 3D interactive visualization and statistical approximation was used to analyse the topographic profiles of the surfaces. Scanning was performed perpendicular to the axis of the cantilever at a rate of typically 1 Hz, with scan areas of 1 μm × 1 μm, 10 μm × 10 μm and 50 μm × 50 μm.
4. Conclusions
The pp–GT films were successfully prepared and characterized. An FTIR analysis confirmed that exposure to an RF plasma field can effectively initiate the polymerization of the monomer. Some functional groups observed in the monomer were retained during the polymerization process, but the C=O stretching vibration (observed in pp–GT spectra) was not present in the monomer spectrum. An increase in RF power resulted in a reduction of the magnitude of the remaining groups, which was attributed to increased fragmentation and, consequently, polymerization of the monomer. RF power can therefore be an effective tool for the fabrication of polymer thin films with tuned properties.
The pp–GT films exhibited different wetting properties, as indicated by the investigation of CA analysis and wetting envelopes. The trends observed were directed by two main phenomena at the solid/liquid interface, namely absorption and spreading, which affect the overall wetting behaviour. This interpretation of the data was based on the assessment of the geometry of the water droplet placed on the film surfaces. The greatest absorption was detected for 10-W pp–GT films while using water as the solvent. Improved stability was observed for films fabricated at higher RF power with the increased hydrophobicity of the polymer surface, from 61.0° (10 W) to 80.7° (75 W). The polymer demonstrated a strong electron donor component and a negligible electron acceptor component and was therefore monopolar in nature. Wetting curves showed that the samples became more hydrophobic as the deposition energy was increased. Chloroform and chlorobenzene were both found to fall within the 0° boundary, indicating that they would completely wet the surface. Solubility results confirmed that the polymer would resist solubilisation from the solvents investigated.
The wettability and compatibility of pp–GT thin films with solvents utilized in the manufacturing of organic electronic devices has been investigated in this paper. From the wettability studies, it is concluded that pp–GT layers are suitable for encapsulation purpose and also for use in a variety of organic electronic devices requiring solution-processed layers.