Atomised spray plasma deposition of hierarchical superhydrophobic nanocomposite surfaces

A R T I C L E I N F O

In this article, we describe an approach which overcomes the aforementioned disadvantages. This comprises the single-step atomised spray plasma deposition (ASPD) of liquid repellent nanocomposite coatings using a low surface energy precursor-nanoparticle slurry (perfluorotributylamine mixed with methacryloyl functionalised silica, zinc oxide, or graphene nanoparticles), which yields hierarchical roughness and mechanical hardness, Scheme 1. The selection of a fluorocarbon precursor provides for both water and oil repellency, and the utilisation of a sub-atmospheric pressure plasma avoids the requirement for expensive carrier gases as well as providing the safe removal of volatile toxic low molecular by-product species [31].
Atomised spray plasma deposition was carried out in an electrodeless, cylindrical, T-shape glass reactor (volume 1117 cm 3 , base pressure of 3 × 10 −3 mbar, and a leak rate better than 2 × 10 −9 mol s −1 ) [35] enclosed in a Faraday cage. The chamber was pumped by a 30 L min −1 two-stage rotary pump (model E2M2, Edwards Vacuum Ltd.) attached to a liquid nitrogen cold trap, and the system pressure monitored by a thermocouple gauge. An L-C impedance matching network was used to minimise the standing wave ratio for power transmitted from a 13.56 MHz radio frequency (RF) power supply to a copper coil (4 mm diameter, 7 turns) located downstream from an atomiser (20 μm diameter median droplet size [36,37], model No. 8700-120, Sono-Tek Corp.), which was driven by a broadband ultrasonic generator (120 kHz, model No. 06-05108, Sono-Tek Corp.). Prior to each deposition, the chamber was scrubbed with detergent, rinsed with propan-2-ol and acetone (+99%, Fisher Scientific Ltd.), and oven dried. Next, a continuous wave air plasma was run at 0.2 mbar pressure and 50 W power for 30 min to remove any remaining trace contaminants from the chamber walls. Ambient temperature deposition was carried out using a 30 W continuous wave plasma in conjunction with atomisation of the solid-liquid slurry into the reaction chamber employing an optimised flow rate of 16 ± 4 × 10 −4 mL s −1 (higher flow rates produce unstable films due to incomplete polymerisation). Upon plasma extinction, the atomiser was switched off and the system was evacuated to base pressure, followed by venting to atmosphere. The chemical stability of the deposited nanocomposite layers towards polar and non-Scheme 1. Atomised spray plasma deposition (ASPD) of perfluorotributylamine-nanoparticle nanocomposite layer.  Colloids and Surfaces A 558 (2018) 192-199 polar solvents was tested by rinsing the samples with a 1:1 v/v mixture of propan-2-ol/cyclohexane for 1 min and air dried. Control experiments showed that in the absence of plasma ignition, the atomiser deposited layers could be readily washed off with polar and non-polar solvents.

Contact angle analysis
Sessile drop static contact angle measurements were carried out at 20°C using a video capture apparatus in combination with a motorised syringe (model VCA 2500XE, A.S.T. Products Inc.). 1.0 μL droplets of ultrahigh-purity water (B.S. 3978 grade 1) and hexadecane (99%, Sigma-Aldrich Ltd.) were employed as probe liquids for hydrophobicity and oleophobicity respectively. Advancing and receding contact angle values were determined by respectively increasing the dispensed 1.0 μL liquid drop volume by a further 1.0 μL, and then decreasing the liquid drop volume by 1.0 μL [38].

X-ray photoelectron spectroscopy
Deposited layers were analysed by X-ray photoelectron spectroscopy (XPS) using a VG ESCALAB II electron spectrometer equipped with a non-monochromated Mg Kα X-ray source (1253.6 eV) and a concentric hemispherical analyser. Photoemitted electrons were collected at a take-off angle of 20°from the substrate normal with electron detection in the constant analyser energy mode (CAE, pass energy = 20 eV) [39]. Experimentally determined instrument sensitivity (multiplication) factors were C(  [39]. A linear background was subtracted from core level spectra and then fitted using Gaussian peak shapes with a constant full-width-half-maximum (FWHM) [40,41]. All binding energies are referenced to the Mg Kα 1,2 C(1s) eCF 2 e peak at 291.2 eV binding energy [42,43].

Infrared spectroscopy
Fourier transform infrared (FTIR) analysis was carried out using an FTIR spectrometer (Spectrum One, Perkin Elmer Inc.) equipped with a liquid nitrogen cooled mercury cadmium telluride (MCT) detector. The spectra were averaged over 285 scans at a resolution of 4 cm −1 across the 450-4000 cm −1 range. Reflection-absorption infrared spectroscopy (RAIRS) of ASPD nanocomposite layer coated silicon wafers was performed using a variable angle reflection-absorption accessory (Specac Ltd.) fitted with mirrors aligned at an angle of 66°to the substrate normal.

Scanning electron microscopy
ASPD coated silicon wafers were mounted onto carbon disks supported by aluminium stubs, and then covered with a 5-10 nm evaporated gold layer (Polaron SEM Coating Unit, Quorum Technologies Ltd.). Surface morphology images were acquired on a scanning electron microscope (model Vega 3LMU, Tescan Orsay Holding, a.s.) operating in secondary electron detection mode at an accelerating voltage of 8 kV, and a working distance of 8-10 mm.

Microindentation
Vickers hardness (HV) values were measured using a micro Vickers hardness tester (model MVK-H2, Mitutoyo Inc.) and then converted into GPa. A standard Vickers indenter tip was employed with applied loads of 98, 245, 490, and 980 mN (international standard test ASTM E384-11e1) [44]. The tip load was applied for 10 s, at an indentation speed of 3 μm s −1 and then unloaded over a period of 10 s. At least 5 different sampling points across the surface were analysed for each applied load value.

Deposition rate
Atomised spray deposition using perfluorotributylamine in the absence of plasma ignition resulted in negligible film growth rate (below 0.1 ± 0.1 nm min −1 following solvent washing of the deposited layer), thereby signifying the importance of plasma activation of the atomised droplets as well as the substrate surface for adhesion. The optimal atomised spray plasma deposition (ASPD) rate for the perfluorotributylamine precursor was measured to be 49 ± 4 nm min −1 at a liquid flow rate of 16 ± 4 × 10 −4 mL s −1 . This value is an order of magnitude greater than that reported for conventional vapour phase perfluorotributylamine plasma deposition (5.9 nm min −1 growth rate [45])-which can be attributed to the higher precursor flow rate for atomised liquid droplets.

Contact angle
The wettability of the optimal deposition rate ASPD perfluorotributylamine layer (water contact angle = 114 ± 1°) was found to be comparable to its conventional vapour phase plasma deposited counterpart (water contact angle = 111°for coated flat substrate [46]), Fig. 2. A level of oleophobicity was also measured (hexadecane contact angle = 65 ± 1°) which is consistent with the reported hexadecane contact angle value of 68°for C8-perfluoroalkyl chain (perfluorooctyltrichlorosilane) self-assembled monolayers on flat silicon surfaces [47].
Incorporation of methacryloyl-SiO 2 nanoparticles into the ASPD perfluorotributylamine layer led to a significant enhancement in liquid repellency yielding water and hexadecane contact angles as high as 168 ± 5°and 90 ± 10°respectively for an optimal precursor slurry loading of 0.75% w/w silica nanoparticles, Fig. 2. For nanoparticle concentrations exceeding this loading, the nanoparticle slurry mixture became too viscous to sustain homogeneous atomisation. These liquidrepellent nanocoatings were stable towards washing with a 1:1 v/v propan-2-ol/cyclohexane polar/non-polar solvent mixture.

X-ray photoelectron spectroscopy
XPS analysis of the ASPD perfluorotributylamine layer detected the presence of only carbon, fluorine, and nitrogen, Table 1. The absence of any Si(2p) and O(1s) XPS signals confirmed pin-hole free coverage of the deposited layer over the underlying silicon substrate. For the case of ASPD perfluorotributylamine-methacryloyl-SiO 2 nanocomposite and perfluorotributylamine-(methacryloyl-SiO 2 + ZnO) nanocomposite layers, less than 0.2 at.% of silicon or zinc XPS signal, and a small amount of oxygen were detected, which confirms that the nanoparticles remain encapsulated within perfluorotributylamine-nanoparticle slurry droplets during atomised spray plasma deposition (0.2-5 nm XPS sampling depth [48]).
The C(1s) XPS spectra of ASPD perfluorotributylamine (and nanocomposite) layers were fitted to five Gaussian Mg Kα 1,2 components in conjunction with their corresponding Mg Kα 3 and Mg Kα 4 satellite peaks shifted towards lower binding energies by ∼8.4 and ∼10.2 eV respectively, Fig. 4 [49]. The C(1s) Mg Kα 1,2 components being:  Table 1. There was no significant variability in the chemical composition or F:C ratio between the various ASPD perfluorotributylamine-nanocomposite layers, Supplementary Material Fig. S1.
The ASPD perfluorotributylamine-graphene nanocomposite layer displayed a strong characteristic graphene infrared absorbance feature in the 600-450 cm −1 region.

Scanning electron microscopy
Scanning electron microscopy (SEM) images of ASPD perfluorotributylamine layers showed a flat surface morphology indicating the deposition of a smooth nanocoating, Fig. 6.
Incorporation of the various types of nanoparticles gave rise to hierarchical topographical structures. ASPD perfluorotributylamine-methacryloyl-SiO 2 nanocomposite layers present dispersed 3-level hierarchical roughness islands comprising a background nanoscale roughness superimposed onto microscale spherical asperities and larger cavities (ca. 12 μm diameter)-which correlate to the enhancement in water and hexadecane contact angle values, Fig. 2. A mixture of methacryloyl-SiO 2 and ZnO nanoparticles in the ASPD nanocomposite layers also resulted in hierarchical roughness but yielded a more evenly distributed hierarchical surface structure (no large-scale cavities which manifests in higher hexadecane contact angle values)-this may arise due to a better dispersed perfluorotributylamine/(methacryloyl-SiO 2 + ZnO) nanoparticle slurry mixture through the use of trifluoroacetic acid fluorosurfactant, Fig. 3. Such hierarchical roughness lowers liquid-solid interaction due to air pockets in accordance with the Cassie-Baxter model for surface wetting [14].
ASPD nanocomposite layers containing graphene lacked significant nanoscale structure and presented a more globular microscale roughness (presumably due to the larger platelet size of graphene), and consequently displays the lowest hexadecane contact angle values amongst the range of ASPD nanocomposite layers, Fig. 3.

Microindentation
Microindentation measurements showed nanoparticle incorporation significantly improves the hardness of ASPD nanocomposite layers. Also, they displayed indentation-resistance at applied loads below 245 mN, Fig. 7. In all cases, the hardness improved by at least two-fold.

Discussion
Atomised spray plasma deposition (ASPD) is a solventless, singlestep, and substrate-independent method for the deposition of functional nanocoatings [66][67][68][69][70][71]. Nanoparticles mixed with a fluorocarbon precursor to form a slurry mixture have been atomised into an electrical discharge and directed towards a target substrate. Plasma-excited species (mainly electrons, ions, and radicals) activate precursor-nanoparticle slurry droplets during impact onto the plasma-activated substrate leading to nanocomposite film growth.
Perfluorocarbon groups display weak intermolecular interactions due to the high electronegativity and electron-withdrawing effect of fluorine atoms. Hence, long perfluorocarbon chain lengths are able to lower the surface energy because of such weak intermolecular forces, thereby enhancing liquid repellency [72]. This accounts for the hydrophobic contact angle measured for the ASPD perfluorotributylamine layer containing no nanoparticles, where a lack of surface roughness is likely to make the Cassie-Baxter effect insignificant, Figs. 3 and 6. The XPS elemental composition N:C:F ratio of 1.0:5.1:8.4 for the ASPD perfluorotributylamine layer can be correlated to the characteristic low energy electron-impact fragmentation molecular ion formed from perfluorotributylamine in the gas phase: C 5 F 10 N + (m/z of 264) with a molecular structure of CF 3 CF 2 CF 2 CF]N + ]CF 2 (N:C:F ratio of 1:5:10), Table 1, and Supplementary Material Table S2 [73,74]. This is consistent with the high level of nitrogen atom incorporation measured by XPS (6.9 at.% compared to the precursor theoretical value of 2.5 at.%, Table 1). The associated unsaturation and crosslinking in the deposited layer gives rise to a hard polymeric nanocoating, Fig. 7.
Further enhancement in liquid repellency has been achieved  through the introduction of micro-/nanoscale hierarchical roughness by incorporating nanoparticles to generate a composite layer in accordance with the Cassie-Baxter model [14] and the lotus leaf effect [13]. This inclusion of different sized nanoparticles within the perfluorocarbon plasma polymer host matrix improves liquid repellency as well as the nanocomposite film mechanical properties, Figs. 3 and 7. By utilising reactive (methacryloyl) functionalised silica nanoparticles, greater bonding is promoted within the growing fluorocarbon polymer matrix via plasma excitation. For comparable applied loads (980 mN), the ASPD perfluorotributylamine-graphene nanocomposite layer displays hardness values exceeding stainless steel (10.7 GPa versus 1-2 GPa respectively) [75].

Conclusions
Atomisation of fluorocarbon precursor-nanoparticle slurries into a low temperature non-equilibrium electrical discharge leads to the deposition of nanocomposite layers. For the case of perfluorotributylamine based slurries containing methacryloyl functionalised silica, zinc oxide, or graphene nanoparticles, it has been found that low surface energy hierarchical roughness nanocoatings are deposited which display enhanced levels of repellency towards water and oil liquids. There is also a significant improvement in mechanical properties yielding an order of magnitude greater hardness compared to stainless steel.

Conflicts of interest
There are no conflicts of interest to declare.