Surface & Coatings Technology 378 (2019) 124929

Contents lists available at ScienceDirect

Surface & Coatings Technology

journal homepage: www.elsevier.com/locate/surfcoat

Understanding the correlations between the mechanical robustness, coating T

structures and surface composition for highly-/super-hydrophobic ceramic
coatings
⁎ ⁎⁎
Pengyun Xu, Thomas W. Coyle , Larry Pershin, Javad Mostaghimi
Centre for Advanced Coating Technologies, University of Toronto, Toronto, ON M5S 3G8, Canada

A R T I C LE I N FO A B S T R A C T

Keywords: Highly-hydrophobic dense Yb2O3 coatings with bump structures and relatively flat topographies were deposited
Solution precursor vacuum plasma spray via the solution precursor vacuum plasma spray (SPVPS) process, and superhydrophobic porous coatings with
Highly-/super-hydrophobic coating distinct columnar structures were deposited via the solution precursor atmospheric plasma spray (SPAPS)
Mechanically-robust process. The SPVPS coating exhibited a water contact angle (WCA) of ~155° and a roll-off angle (RA) of ~28°,
Surface structure
while the SPAPS coating exhibited a WCA of ~163° and RA of ~3°. The SPVPS coatings maintained high mi-
Hydrocarbon
crostructure integrity and relatively consistent hydrophobicity after a comprehensive set of mechanical tests,
showing a much higher mechanical robustness. In contrast, the columnar structures of the SPAPS coatings were
severely damaged by mechanical contacts, resulting in a sharp WCA decrease. The WCA of the abraded coatings
increased after vacuum treatment due to hydrocarbon re-adsorption; the abraded SPAPS coating regained its
initial superhydrophobicity while the SPVPS coating showed a mild WCA increase. The surface structures and
compositions of the original coatings, the abraded coatings and vacuum-treated coatings were investigated,
revealing close correlation of the changes in WCA with the surface composition and structures.

1. Introduction non-wetting surfaces. Improving the mechanical robustness of non-

wetting surfaces is of great significance to elevate the durability of the
Non-wetting surfaces exhibiting high water contact angles (WCA) hydrophobicity and move the non-wetting surfaces towards practical
have received much attention due to their potential for use in self- applications.
cleaning, anti-icing, oil/water separation, drag reduction, and other A large number of processes have been used to attempt to fabricate
applications [1]. A surface presenting a WCA larger than 150° and a mechanically robust non-wetting surfaces. Surface layers with self-si-
roll-off angle (RA) smaller than 10° is referred to as a superhydrophobic milar microstructure, texture and functionality extending for some
surface. Superhydrophobic surfaces are generally combinations of low- depth below the surface were fabricated so that similar wetting beha-
surface-energy materials and hierarchical surface structures [2]. De- viors were maintained even when the top layer was abraded or da-
spite the excellent water repellency of the artificial non-wetting sur- maged [5,6]. Some non-wetting surfaces exhibited good abrasion re-
faces, the lack of mechanical robustness has severely restricted their sistance due to the hard, mechanically robust nature of the surface
practical applications [3]. The micron/nano-scale surface topography, material, such as metallic coatings modified by low-surface-energy
which provides the typical hierarchical structure maintaining a super- material [7] and composite ceramic coatings [8]. Some fabrication
hydrophobic Cassie-Baxter state [4], can be easily damaged or removed processes introduced protective sacrificial micro-pillars into the surface
if the structures are not sufficiently mechanically-robust. The non- [9] or created multi-level structures [10] to protect the fine-scale sur-
wettability of surfaces is consequently degraded due to damage or loss face roughness during the abrasion process. Coatings with high flex-
of the hierarchical roughness and the change of chemical composition ibility and compliance have demonstrated good mechanical robustness
of the surfaces. Mechanical contacts such as rubbing, abrasion, scrat- via elastic deformation, such as the copper-based coatings with coral-
ches and erosion are very common circumstances for applications of reef surface structures [11] and all-organic nanocomposite coatings [6].

⁎
Correspondence to: T.W. Coyle, Department of Materials Science and Engineering, University of Toronto, 184 College Road, Toronto, Ontario M5S 3E4, Canada.
⁎⁎
Correspondence to: J. Mostaghimi, Department of Mechanical & Industrial Engineering, University of Toronto, 5 King's College Road, Toronto, Ontario M5S 3G8,
Canada.
E-mail addresses: tom.coyle@utoronto.ca (T.W. Coyle), mostag@mie.utoronto.ca (J. Mostaghimi).

https://doi.org/10.1016/j.surfcoat.2019.124929
Received 11 July 2019; Received in revised form 21 August 2019; Accepted 22 August 2019
Available online 23 August 2019
0257-8972/ © 2019 Elsevier B.V. All rights reserved.
P. Xu, et al. Surface & Coatings Technology 378 (2019) 124929

The fabrication of self-healing or easily-repairable superhydrophobic 2. Experimental procedures

surfaces has also drawn significant attention [12,13]. However, most of
the above-mentioned processes are either time-consuming or are not 2.1. Materials and coating deposition
suitable for large-scale fabrication. Developing a fast and simple process
to fabricate mechanically robust non-wetting surfaces is of vital im- As reported previously [30,31], coatings were deposited on sand-
portance to promote practical applications. blasted stainless steel disks using a 1 mol/l Yb(NO3)3 solution via the
The lanthanide rare earth oxides (REOs) were reported to be in- SPVPS process (150 mbar) and SPAPS process (atmospheric pressure). A
trinsically hydrophobic and the hydrophobicity maintained after commercial F4-VB torch (Oerlikon Metco) was operated at a current of
abrasion-wear testing and high-temperature exposure [14]. The lan- 700 A with a plasma gas mixture of 50 LPM Ar and 10 LPM H2 in both
thanide REOs show great potential to be used for the fabrication of processes, which generated a total power of ~45 kW. As the plasma jet
mechanically-robust non-wetting ceramic surfaces, although the basis was extended significantly at low pressure, the standoff distance (SD,
of the hydrophobicity of REOs has been in debate. Many researchers distance between the torch nozzle exit and the substrate surface) in the
claimed the REOs were intrinsically hydrophilic and became hydro- SPVPS process was set as 80, 90, 100 and 110 mm to avoid overheating
phobic because of the adsorption of airborne hydrocarbons on the of the samples, while the SD was set as 30 mm in the SPAPS process. A
surfaces, as demonstrated by the behavior of sintered REO pellets [15], detailed description of the two deposition processes is reported in
pulsed laser-deposited REO films [16,17], and magnetron sputtered Section S1 of the Supplementary material.
REO films [18]. Other publications argue that REOs are intrinsically
hydrophobic as long as the surface reaches a stoichiometric oxygen-to- 2.2. Coating characterizations
metal ratio [19,20]. Another article relates the hydrophobicity to the
low electronegativity of the metal atoms in REOs [21]. Consistent with The crystallographic phase of the as-deposited coatings was ana-
much of the literature [15–18], our previous work [22] also showed lyzed by X-ray diffraction (XRD, MiniFlex 600, Rigaku, USA) using CuKα
hydrocarbon adsorption on the REO surface during a vacuum treatment X-rays from 15° to 105° 2-theta with a scan speed of 1.5°/min. The
process resulted in the hydrophobic behaviors. microstructures of coatings were observed using a scanning electron
Despite the lack of consensus on the nature of the hydrophobicity, microscope (SEM, Hitachi TM3000, Japan). The surface topography
many processes for fabricating hydrophobic REOs surfaces have been and roughness of coatings were measured by a digital optical micro-
reported, including pressing and sintering [15], electrodeposition scope (VHX-5000, Keyence Canada Inc., Canada) and analyzed by
[23–25], pulsed laser deposition [16,17], hydrothermal synthesis [26], Gwyddion software. Surface chemical compositions of the coated sur-
sputtering [18,27], and thermal spraying [20,28]. Many of these pro- faces were analyzed using X-ray photoelectron spectroscopy (XPS, K-
cesses are multi-step, time-consuming and not suitable for large-scale Alpha, Thermo Scientific, USA) with an monochromated AlKα X-ray
fabrication [14,15,26]. Some processes are limited by the substrate (spot size of 400 μm).
dimensions [14,15] or require special pre-treatment of the substrate
surface [25]. In addition, many reported REO surfaces are not super- 2.3. Wettability characterization
hydrophobic (90° < WCA < 150° or RA > 10°) [20,24,27,28], and
so not suitable for certain applications such as self-cleaning. The as-sprayed coatings were superhydrophilic and required a va-
We have previously described the fabrication of superhydrophobic cuum treatment process to become hydrophobic. Our previous work
REO coatings by solution precursor plasma spray (SPPS), which shows showed that hydrocarbon species were adsorbed on the Yb2O3 surface
high feasibility for quick, efficient, large-scale deposition [29,30]. during vacuum treatment, forming a low-surface-energy film [22].
Manipulation of SPPS process parameters allows control of coating Therefore, the as-sprayed coatings were placed in a chamber (1–15 Pa)
structure at the sub-micron or nanometer scale. Following our work on backed by an oil-free diaphragm pump and a turbomolecular pump at
solution precursor atmospheric plasma spray (SPAPS) [29,30], a novel room temperature (~23 °C) for 12 h. The WCA and RA were measured
solution precursor vacuum plasma spray (SPVPS) process was devel- using a sessile drop method and tilting plate method respectively, using
oped by combining the solution precursors with vacuum plasma spray water droplets with a volume of ~7 μl. The WCA and RA were reported
(VPS) [22,31]. VPS refers to a plasma spray process conducted at low as the average of at least five measurements on different areas of the
pressure (50–200 mbar) and is well known to produce denser coatings coatings. WCA and RA are the most common measurements to evaluate
than atmospheric plasma spray (APS) [32]. However, the cost of VPS is the hydrophobicity of a non-wetting surface, however they do not
higher than APS and the size of the structure that can be coated is characterize the durability of the hydrophobicity. The hydrophobicity
limited by the size of the vacuum chamber [33]. The formation of of non-wetting surfaces may be degraded under harsh environments
submicron-/nano-sized particles and splats in both SPAPS and SPVPS such as high-velocity water jetting or flushing. Therefore, dynamic
contributes significantly to the formation of hierarchical surface water impact tests were conducted on the coated surfaces. Water dro-
structures, which are beneficial for achieving the superhydrophobic plets (~7 μl) were released from a height of 100 mm, fell by gravity and
state. impacted five different areas of the coatings. The water impact and
A wide range of REO coating structures and surface topographies rebound process was recorded using a high speed camera at 2000 fps.
can be produced by manipulating the deposition conditions employed As the SPVPS coatings deposited at different SD showed very similar
in SPAPS and SPVPS, resulting in different wetting behaviors. The microstructures and wetting behaviors (reported later), only the SPVPS
surface structures and the amount of hydrocarbon adsorbed on the coating of SD = 80 mm was used for the comparison with the SPAPS
coatings could be changed by mechanical contact. In the following we coating (SD = 30 mm) in the wettability characterization and me-
examine how the mechanical robustness of dense, highly-hydrophobic chanical tests.
Yb2O3 coatings deposited via the SPVPS process and columnar, super-
hydrophobic coatings deposited by the SPAPS process is affected by 2.4. Mechanical tests
differences in the coating structure through a comprehensive set of tests
including scratch, tape peeling, cross-cut, erosion and linear abrasion. As no standardized characterization method has been established to
The effects of abrasion and vacuum treatment on the wetting behaviors evaluate the mechanical robustness of non-wetting surfaces, several
of the coatings were investigated by analyzing the coating micro- types of tests were conducted on the vacuum-treated coatings following
structures, surface topographies, and surface compositions before and procedures reported in the literature [34–36]. Mechanical tests in-
after the mechanical tests. cluding scratch, tape peeling, cross-cut, erosion and linear abrasion
were performed until an obvious WCA decrease or severe coating

2
P. Xu, et al. Surface & Coatings Technology 378 (2019) 124929

damage was observed. The coated surfaces were cleaned using com- shown in Fig. 2f, with the clusters and inter-cluster gaps shown as red
pressed air after each mechanical test followed by immediate char- islands and blue valleys respectively. The average roughness of the
acterization of wetting behaviors and microstructures. coating measured from the topographical contour was ~12.6 μm.
The adhesion of the coatings to the substrates was characterized by In contrast to the SPAPS coatings, the SPVPS coatings exhibited a
the scratch test, tape peeling test and cross-cut test. The scratch test was distinctly different type of microstructure due to the different char-
conducted using a custom-made scratch tester based on a Vickers acteristics of plasma under low pressure and atmospheric pressure, as
hardness tester (ZwickRoell, Germany). A conical diamond stylus (cone shown in Fig. 2b–e. The microstructures did not change significantly
apex angle of 90°, tip radius 200 μm) was used to scratch five different with the standoff distance, which was attributed to the uniform heating
areas of each coating at a speed of ~2 mm/s under different loads for a of the Yb2O3 droplets and small decay rate of gas temperature and
distance of 5 mm. The tape peeling test was conducted following the velocity of the jet with distance. The microstructures were much denser
procedure reported in literature [34,37]. In brief, a tape (Elcometer 99 (porosity 7%–9%) than the SPAPS coatings, which was attributed to the
ASTM tape, adhesion to stainless steel of 668 N/m) was applied on the supersonic gas flow and higher droplet velocity. Different than the
coated surface, pressed by a 500 g stainless steel roller, and then peeled columnar structures formed in the SPAPS coatings, structures char-
off at a rate of ~2 mm/s after a dwell time of 90 s. The cross-cut test was acterized by very short bumps (20–30 μm in diameter, 5–8 μm in
performed following the procedure reported in ASTM D3359-02 stan- height), without obvious inter-bump gaps were formed on the coating
dard. Depending on the thickness of the coatings, cross-hatch patterns surfaces. The short bump structures of the coating deposited at 80 mm
(11 cuts with a 1 mm spacing for SPVPS coatings; 6 cuts with a 2 mm are shown as small red islands in the topographical contour (Fig. 2g),
spacing for SPAPS coatings) were made on the coatings, followed by a and the average roughness was measured to be ~3.4 μm. The SPVPS
tape (Elcometer 99 ASTM tape) peeling test on the patterns. Three tape coatings deposited at all standoff distances had an average roughness in
peeling tests and cross-cut tests were conducted on each coating to the range of 3.2 to 3.5 μm due to the similar surface structures.
ensure the repeatability. These SPVPS coatings were also distinctly different than the pre-
The erosion test was conducted by blasting the coated surfaces using viously-reported SPVPS coatings [22]. The orthogonal deposition in this
a sandblast machine, which was different than previously reported tests work produced dense microstructures and relatively flat surface topo-
in which the particles were poured from a certain height to impact the graphies as shown in Fig. S2a. In contrast, the glancing angle deposition
surface by gravity [38–40]. Alumina grit (average size of 102 μm) im- method applied in the previous work resulted in cauliflower-like clus-
pinged the surface at an angle of 45° under a pressure of 20 Psi from a ters with inter-cluster gaps (see Fig. S2b) [22]. Therefore, even though
distance of 200 mm. The feeding rate of alumina grit was 300–350 g/ very fine particles and splats were formed on the surface, the SPVPS
min and the impingement velocity was ~6 m/s. The area impacted by coating did not show the same hierarchical structure as the SPAPS
the grit covered the entire coated surface. A 5-second sandblast was coatings or the previously-reported SPVPS coating [22]. Little air could
defined as one erosion cycle and the impinging energy of the grit in one be trapped in the inter-bump gaps, which made forming a stable Cassie-
erosion cycle was calculated to be 1–1.2 μJ. Linear abrasion tests were Baxter state difficult [4].
performed by moving the coating reciprocally against grinding paper
(1200 grit) at a speed of ~100 mm/s under a constant pressure of 3.2. Wetting behaviors of the coatings
2.2 kPa. Abrasion over a length of 200 mm was defined as one cycle.
The sample was turned 90° and moved against a new area of the The as-deposited SPVPS coatings and SPAPS coatings both exhibited
sandpaper after every cycle to obtain uniform abrasion over the coated high water repellency after vacuum treatment, as shown in Fig. 3.
surface. Five abrasion tests were performed on each kind of coating to Water droplets of different volumes sitting on the coated surfaces ex-
ensure repeatability. hibited near-spherical morphologies. The SPVPS coating presented a
WCA of ~155° and a much larger RA of ~28° than the SPAPS coating
3. Results and discussion (WCA of ~163°, RA of ~3°). Water droplets rolling off the SPAPS
coating exhibited an almost spherical morphology, indicating a low
3.1. Phase composition, microstructure and surface topography of the resistance to rolling and the high mobility of water droplet. In contrast,
coatings the water droplets adhered to the SPVPS coating even when tilted
significantly before rolling, indicating the coating was more adhesive to
The XRD patterns of the as-sprayed SPAPS and SPVPS coatings de- water.
posited at different standoff distances (SD) are shown in Fig. 1. The For the SPVPS coatings, the bump structures were very short and the
sharp peaks of all the coatings matched the Yb2O3 reference pattern surface was relatively flat without obvious inter-bump gaps. The wet-
(01-077-0455) at all peak locations, indicating the transition from Yb ting state of water on the surface of SPVPS coatings was close to the
(NO3)3 precursor to highly crystalline cubic Yb2O3 coating during both Cassie impregnating state [42], where water impregnated the shallow
deposition processes. inter-bump gaps while the roughness created by the nano-sized parti-
The cross-sectional and top surface microstructures of the SPAPS cles/splats was unwetted resulting in a large WCA. However, the large
coatings are shown in Fig. 2a. The coatings showed columnar structures solid-liquid contact area fraction created a higher adhesion to the water
formed from micron-sized clusters with inter-cluster gaps. The forma- droplet, resulting in a high RA. The very large WCA and effortless
tion of the columnar structures was attributed to a shadowing effect as rolling-off behavior of water droplets on the SPAPS coating indicated
described in our previous work [30], wherein small particles entrained the coating stayed in Cassie-Baxter state due to the hierarchical co-
in the jet might readily follow the diverted plasma gas streams parallel lumnar structures and air trapped in the inter-cluster gaps [4].
to the substrate surface and impinge on surface asperities rather than The droplet behaviors in the dynamic water impact tests were very
impacting the surface orthogonally. The coatings exhibited quite porous different for the two coatings. As shown in Fig. 3c, the water droplet
microstructures (porosity 34%–40% within a column structure) as stretched upwards severely after the impact on the SPVPS coating, re-
shown in the high-magnification insets of the cross-sectional micro- sulting in the detachment of several small droplets. However, the dro-
structure in Fig. 2a. Numerous fine particles, splats and pores were plet could not completely rebound, and a large portion of the droplet
observed in the cross sections and on the top surfaces of the coatings. adhered to the coated surface. When the droplet became stationary
The nano-sized particles and splats superimposed on the micron-sized after the merging of the detached small droplets and the remaining
clusters generated hierarchical structures that have been shown to be droplet, the WCA was measured to be ~125° and the water droplet did
beneficial for achieving a stable Cassie-Baxter state and generating the not roll off even when the surface was vertical or inverted. This in-
lotus effect [4,41]. The topographical contour of the SPAPS coating is dicated the wetting state transitioned from the Cassie impregnating

3
P. Xu, et al. Surface & Coatings Technology 378 (2019) 124929

Fig. 1. XRD patterns of the SPVPS and SPAPS coatings.

state to the Wenzel state under the effect of the high-velocity water peeling cycles. The RA of the SPVPS coating increased quickly to ~75°
impact [42,43]. For the SPAPS coating (Fig. 3d), the water droplet after only two peeling cycles, and the water droplet could not roll-off
exhibited multiple complete rebounding behavior without any water after the third peeling cycle. In spite of the hydrophobicity degradation,
remaining on the surface. The WCA and RA measured after the droplet no obvious damage of the top surface microstructure of the SPVPS
became stationary were ~158° and 6° respectively, indicating the coating was observed (Fig. 5b). The microstructure of the coating re-
coating was still superhydrophobic after the high-velocity water im- mained intact and undamaged after ten more peeling cycles, and the
pact. WCA decreased to ~135°.
In contrast to the intact SPVPS coating, the peel-off test resulted in
3.3. Microstructures and wettability of coating after mechanical tests severe delamination of the SPAPS coating. Many areas of bare substrate
surface were observed after five peeling cycles. The microstructures of
The microstructures and wettability of the SPVPS coating and the residual coated area are shown in Fig. 5c. About 10% of the co-
SPAPS coating after different mechanical tests are reported in the fol- lumnar structures were broken and removed after one peeling cycle,
lowing. The scratch width of the coatings after scratch testing was re- and over 60% of the columnar structures were broken after five peeling
corded by averaging the scratch width at five different areas using cycles, leaving residual parts of columnar structures on the surface, as
500× SEM images and plotted as a function of the load applied in shown in Figs. 5c and S3. Although the coating was severely delami-
Fig. 4. The scratch width of the SPVPS coating increased more slowly nated after the peel-off test, the WCA and RA of the SPAPS coating
with load than of the SPAPS coating. The scratch width of SPVPS could be measured on the remaining coated areas. In these areas the
coating was over two times smaller than that of SPAPS coating under WCA decreased to ~150° after 5 peeling cycles, and the RA increased
the same load, indicating a much higher hardness. The micrographs gradually to ~10°. No further peel-off tests were performed on the
inset in the graph show that the micron-sized clusters of the SPAPS SPAPS coating due to the severe coating delamination.
coating were crushed by the indenter at 200 g resulting in flat tracks The top surface microstructures of coatings after the cross-cut tests
with coating debris, while the SPVPS coating showed much better are shown in Fig. 6. For the SPVPS coating, only a small amount of
scratch resistance with narrower scratch tracks. Partial coating de- coating delaminated along the incisions of the cross hatch squares (see
tachment was observed when the load increased to 350 g for the SPAPS Fig. 6a1), while the coating inside the squares remained intact (see
coating, while the coating detachment occurred when the load in- Fig. 6a2), indicating a strong adhesion. The average damage area of
creased to 700 g for the SPVPS coating. each square measured using ImageJ software was smaller than 5%,
The wetting behaviors and top surface microstructures of coatings which was rated as 4B classification according to the ASTM D3359–02
after tape peel-off tests are shown in Fig. 5. The WCA of the SPVPS standard. In contrast, severe delamination was observed on the SPAPS
coating dropped from ~155° for the original coating to ~141° after 5 coating after the cross-cut test as shown in Fig. 6b. The columnar

4
P. Xu, et al. Surface & Coatings Technology 378 (2019) 124929

Fig. 2. Microstructures and topographical contours of the SPAPS coating and SPVPS coatings deposited at different standoff distances.

structures were largely removed along the incisions (see Fig. 6b1), and while the value remained relatively constant (~140°) in the further four
over 50% of the columnar structures inside the squares were broken erosion cycles. The RA increased sharply to 55° after one erosion cycle
(see Fig. 6b2) due to the tape peeling, leaving the residual parts on the and the water droplet could not roll off after two erosion cycles. The
substrate. The SPAPS coating did not detach completely from the sub- impacting alumina grits eroded the fine particles and splats on the
strate after the cross-cut test, and most of the damage was due to coating, resulting in smooth erosion facets on the bump structures as
fracture of individual columnar structure. This indicated a very poor shown in Fig. 7b. Different than the SPVPS coating, the WCA and RA of
cohesion strength for the columnar structures due to the porous mi- the SPAPS coating remained constant after the first erosion cycle
crostructures, as evidenced by the porous fracture surface in Fig. 6b2. (Fig. 7a), although ~12% of the columnar structures were removed
The average coating damage area was measured using ImageJ software, (Fig. S4). The WCA decreased sharply to ~96° after five erosion cycles
showing the SPAPS coating belonged to the 1B classification according and all the columnar structures were broken. Areas of bare substrate
to the ASTM D3359-02 standard. surfaces were observed after six erosion cycles. The RA increased to ~7°
The wetting behaviors and top surface microstructures of coatings after two erosion cycles, and the water droplet could not roll off after
after erosion tests are shown in Fig. 7. The WCA of the SPVPS coating three erosion cycles.
decreased dramatically from ~155° to ~142° after one erosion cycle, It is significant that the WCAs of the eroded coatings increased after

5
P. Xu, et al. Surface & Coatings Technology 378 (2019) 124929

Fig. 3. The wetting behaviors of the coatings. The scale bar in each figure is 2 mm.

Fig. 4. The scratch width under different loads. Insets are the top surface microstructures of the scratch. The scale bar is 100 μm.

6
P. Xu, et al. Surface & Coatings Technology 378 (2019) 124929

Fig. 5. Wetting behaviors and top surface microstructures of coatings after tape peel-off tests. The WCA and RA for the SPAPS coating were measured on areas where
the coating was not delaminated. The scale bar is 50 μm.

Fig. 6. Top surface microstructures of coatings after cross-cut tests. (a2) and (b2) are the high-magnification images of the rectangular area in (a1) and (b1)
respectively.

7
P. Xu, et al. Surface & Coatings Technology 378 (2019) 124929

Fig. 7. Wetting behaviors and top surface microstructures of coatings after erosion tests. The scale bar is 50 μm.

a 12-hour vacuum (1–15 Pa) treatment process, as pointed out by the abrasion are shown in Fig. 8c1. In contrast to the original coating
green arrows in Fig. 7. The WCA of the SPAPS coating increased from (Fig. 2a1), the clusters were severely abraded, and most were broken
~96° after five erosion cycles to ~155° after vacuum treatment, and the and lost from the surface. The residual clusters still exhibited very
RA decreased to ~5°, indicating the coating regained super- rough profiles, and overall a structure of short columns separated by
hydrophobicity. In contrast, the WCA of the SPVPS coating only in- gaps remained (Fig. 8c1). Numerous fine particles, splats and pores
creased from ~137° after five erosion cycles to ~141° after vacuum were observed on the surface of residual clusters, as shown in the inset
treatment. The variation of the wetting behaviors after vacuum treat- of Fig. 8a2. Different than the severe damage of the SPAPS coating, the
ment is discussed in Section 3.4. taller bump structures of the SPVPS coating behaved like protective
The microstructures of the coatings after linear abrasion tests are pillars and protected the microstructure below the bumps (referred to
shown in Fig. 8. The SPAPS coating showed such poor abrasion re- as base microstructures in the following discussion) from abrasion. As
sistance that even one cycle of abrasion left an obvious track of cluster shown in Fig. 8c2, only the tops of the bump structures were abraded
removal, as shown in Fig. 8a1. In comparison, one cycle of abrasion on after 150 cycles of abrasion, while the remaining areas retained the
a SPVPS coating left only a very slight mark on the top of the bump original microstructure and morphologies. This contributed sig-
structures (Fig. S5). A high-magnification image of the microstructure nificantly to maintaining relatively consistent wettability until the taller
of the broken SPAPS clusters (Fig. S6) revealed the very porous nature bump structures were abraded to similar heights with the base micro-
of the clusters, which resulted in the weak abrasion resistance. Most structures. With further abrasion to 500 cycles, more bump structures
clusters were removed after 5 cycles of abrasion and almost no clusters were abraded to the height of the base microstructures, as shown in
remained after 25 cycles, as shown in Fig. 8a2. In contrast to the SPAPS Fig. 8c2.
coating, the SPVPS coating showed a much higher abrasion resistance; The topographical contour of the SPAPS coating after abrasion is
only a limited number of the bump structures were abraded after shown in Fig. 9a. The clusters shown as large red islands in the topo-
150 cycles of abrasion, as shown in Fig. 8b1. The abraded bump graphical contour of the original coating changed to small pieces cor-
structures exhibited relatively smooth and flat plateaus, as a result of responding to the residual parts of the clusters after abrasion. The
the denser microstructures formed in SPVPS process. The area fraction coating abrasion was accompanied by a sharp decrease of the coating
of the abraded surfaces increased with abrasion, while a majority of thickness and average roughness. The thickness (~66 μm) and average
surface remained the original microstructures even after 500 cycles of roughness (~12.6 μm) of the original coating decreased to ~25 μm and
abrasion (Fig. 8b2). The abraded top surfaces of the SPVPS coatings 3.4 μm respectively after 25 cycles of abrasion. The topographical
deposited at other standoff distances showed very similar micro- contour of the original SPVPS coating showed small red islands corre-
structures after abrasion, as shown in Fig. S7. sponding to the bump structures. As a large portion of bump structures
The cross-sectional microstructures of the SPAPS coating after were abraded to plateaus, the surface became much smoother and the

8
P. Xu, et al. Surface & Coatings Technology 378 (2019) 124929

Fig. 8. Microstructures of the coatings after linear abrasion tests: (a1–a2) top surfaces of SPAPS coating; (b1–b2) top surfaces of SPVPS coating; (c1) cross section of
SPAPS coating; (c2) cross section of SPVPS coating.

contour showed relatively uniform red color after 500 cycles of abra- mechanisms. The surface profile of the original SPAPS coating showing
sion (see Fig. 9b). The average roughness decreased from ~3.4 μm for similar cluster height changed to a rough profile corresponding to the
the original coating to ~2.2 μm after 500 cycles of abrasion. residual cluster parts which resulted from a random break-off of the
The abrasion mechanisms of the SPVPS and SPAPS coatings are columnar structures. In contrast, the bump structures of the SPVPS
different. The clusters of the SPAPS coating had very low mechanical coating were abraded gradually and resulted in a relatively flat surface
strength due to the porous microstructures. When the coating was profile after abrasion.
moved against the grinding paper, the tangential force of abrasion in- The WCA variations of the SPVPS coating and SPAPS coating with
itiated cracks which propagated quickly through the connections of abrasion are shown in Fig. 9d. The WCA of the SPVPS coating remained
pores and voids. Therefore clusters easily broke off below the surface of relatively constant, decreasing from ~155° for the original coating to
the coating during abrasion resulting in rapid damage to the surface ~140° after 250 cycles of abrasion. The WCA showed a faster decrease
structures. The abrasion mechanism was characterized by a random from 250 to 350 cycles of abrasion, and decreased to ~107° after
break-off of clusters, with the residual clusters exhibiting very rough 500 cycles of abrasion. The standard deviation of WCA (shown as the
surfaces, as shown in Fig. 8c1. In contrast, the dense bump structures of error bar in Fig. 9d) also increased with abrasion, indicating WCAs
the SPVPS coatings had much higher mechanical strength, so that the became less uniform over the surface after abrasion. The WCA of the
top surfaces of bump structures were abraded gradually, as evidenced SPAPS coating decreased sharply after abrasion, from ~132° to ~100°
by the smooth and dense abraded surfaces of the bump structures after 5 and 25 cycles of abrasion respectively. The RA changed more
(Fig. 8b). This mitigated the removal of coating material and reduced rapidly with abrasion for both coatings. Water droplets could not roll
the coating abrasion significantly. The surface profiles of the coatings off the SPAPS coating after only 5 cycles of abrasion, while the RA of
before and after abrasion (Fig. 9c) verified the different abrasion the SPVPS coating increased to ~38° after 5 cycles of abrasion and

9
P. Xu, et al. Surface & Coatings Technology 378 (2019) 124929

Fig. 9. Topographical contours, surface profiles and wetting behaviors of the coatings after linear abrasion tests: (a) topography of SPAPS coating; (b) topography of
SPVPS coating; (c) surface profiles; (d) wetting behaviors.

water droplets adhered to the coating after 25 cycles. SPVPS coating and SPAPS coating after vacuum treatment were named
As observed for the WCA variation of the eroded coatings, the WCA as SPVPS VAC and SPAPS VAC respectively.
of coatings after linear abrasion test also increased after the vacuum The survey spectra and the constituent peaks of the C1s spectra of
treatment process, as indicated by the green arrows in Fig. 9d. The WCA several coatings are shown in Fig. 10. The survey spectra showed that
of the SPAPS coating increased from ~100° after 25 cycles of abrasion the C1s intensity relative to Yb intensity was the highest for the original
to ~161° after vacuum treatment, very close to the WCA of the original coatings, while it decreased dramatically after abrasion and increased
coating. In contrast, the WCA of the SPVPS coating increased from significantly after vacuum treatment. The C1s spectra were deconvo-
~107° after 500 cycles of abrasion to only ~126° after vacuum treat- luted into high-resolution constituent peaks in terms of the binding
ment. The WCA decrease after mechanical tests and increase after va- energy. Gaussian shapes were assumed for the constituent peaks, with a
cuum treatment were investigated by correlating with the variations of similar full width at half maximum. The fitted spectra (yellow dash dot
surface compositions and structures in Section 3.4. lines) matched well with the pristine spectra shown as black solid lines.
The most dominant peak of C1s (C1s A in Fig. 10) with a binding energy
of 284.7 eV corresponds to adventitious carbon containing C-C(H)
3.4. Variation of surface chemical compositions
bonds. The adventitious carbon, due to the adsorption of hydrocarbon
species, served as non-polar hydrophobic sites [15,44]. The other
We have shown in previous work that the as-sprayed coating was
constituent peaks of C1s (C1s B and C) with higher binding energy
superhydrophilic and became superhydrophobic after vacuum treat-
correspond to CeO and C-O-C bonds in alcohol groups or ether groups,
ment due to hydrocarbon adsorption on the coated surface [22]. The
and O-C=O bonds in carboxyl groups [44], and are referred to as non
low-surface-energy hydrocarbon species adsorbed on the hierarchically-
C-C(H) in the following discussion. The O1s spectra of the coatings were
structured surface during the vacuum treatment, resulting in super-
deconvoluted and shown in Fig. S8 in the Supplementary material.
hydrophobic behavior. The amount of hydrocarbon adsorbed on the
The atomic percentage (at.%) and atomic ratios of different ele-
coated surface was expected to decrease with the removal of coating
ments and species are shown in Fig. 11. The SPVPS coating had a higher
material in the mechanical tests and increase in the vacuum treatment
at.% of C1s and C-C(H) than the SPAPS coating. The at.% of C1s and C-
process by a re-adsorption process; XPS analyses were conducted to
C(H) decreased after abrasion for the SPVPS coating and SPAPS coating,
confirm the expectations.
while the at.% of oxygen and ytterbium both increased. The changes
The SPAPS coating was damaged quickly and severely by tape
can be seen more clearly in Fig. 11b, where the at.% ratio of C/Yb and
peeling and erosion, showing areas of bare substrate surfaces after only
C-C(H)/Yb decreased for both coatings. This indicated the abrasion
a few cycles of test, not suitable for further investigation of surface
removed the hydrocarbon adsorbed on the surface via the removal of
compositions. Therefore, coatings after linear abrasion test were se-
coating material. The at.% of C1s and C-C(H) of the SPVPS coating had
lected to investigate the variation of surface compositions after me-
a decrease of 6.0% and 7.4% respectively after 500 cycles of abrasion,
chanical tests and vacuum treatment using XPS. The time interval be-
while the at.% of C1s and C-C(H) of the SPAPS coating decreased 11.7%
tween the end of each treatment (linear abrasion or vacuum treatment)
and 13.1% respectively after only 25 cycles of abrasion, about two
and the beginning of XPS analysis was less than 30 min to minimize
times more. This was due to the much higher abrasion resistance of the
possible contamination by airborne hydrocarbons. The coatings for XPS
SPVPS coating whereby a large portion of coated surface remained
investigations were denoted as follows. The original highly-hydro-
undamaged after abrasion and so the hydrocarbons adsorbed on that
phobic SPVPS coating (SD = 80 mm) and superhydrophobic SPAPS
portion of surface remained. In contrast, the abrasion broke and re-
coating (SD = 30 mm) are referred to as SPVPS coating and SPAPS
moved all the clusters after only 25 cycles of abrasion for the SPAPS
coating respectively. The SPVPS coating after 500 cycles of linear
coating. The hydrocarbons adsorbed on the cluster surfaces were also
abrasion and the SPAPS coating after 25 cycles of abrasion are referred
removed, resulting in a significant reduction in the amount of
to as SPVPS Abrasion and SPAPS Abrasion respectively. The abraded

10
P. Xu, et al. Surface & Coatings Technology 378 (2019) 124929

Fig. 10. Survey spectra and constituent peaks of C1s spectra of different coatings.

hydrocarbons. SPVPS coating, the taller bump structures were abraded gradually in
The at.% of C1s and C-C(H) of the abraded coatings increased sig- the abrasion test creating smooth plateaus and reducing the amount of
nificantly after the vacuum treatment, to levels very close to the ori- hydrocarbon adsorbed on the surface (see Fig. 12a1–a3). The WCA
ginal coatings, as shown in Fig. 11a. The at.% ratio of C/Yb and C-C(H)/ decreased from ~155° to ~107° after 500 cycles of abrasion. After va-
Yb increased after the vacuum treatment, as shown in Fig. 11b. The cuum treatment the WCA of the abraded coating increased to ~126°
increase of carbon content showed that hydrocarbon re-adsorbed on the due to the re-adsorption of hydrocarbon on the surface (see Fig. 12a4).
abraded coatings in the vacuum treatment process. The non C-C(H) The at.% of hydrocarbon increased to 51.5%, very close to the at.% of
species occupied a much smaller fraction of the total carbon content hydrocarbon of the original SPVPS coating (51.6%). However, since the
than the C-C(H) species. Fig. 11b indicates that the at.% ratio of non C- abraded surface was much smoother, air could not be trapped under the
C(H)/Yb did not change significantly with abrasion or subsequent va- water droplet, resulting in a Wenzel wetting state [43]. Therefore, even
cuum treatment, suggesting that the increase in C content was due though hydrocarbon re-adsorbed on the coating surface, the WCA did
primarily to the adsorption of C-C(H) species. not increase significantly.
For the SPAPS coating, the columnar structures were quickly broken
and removed by abrasion, resulting in rough residual clusters and a
3.5. Discussion
dramatic reduction (a 13.1% decrease) of the hydrocarbon amount. The
significant change in surface structure and reduction in the amount of
The change in wetting behavior of the coatings after linear abrasion
hydrocarbon resulted in the collapse of the Cassie-Baxter state and a
depends on both the change in surface topography and the change in
very sharp WCA decrease from ~163° to ~100° after 25 cycles of
surface composition. The variations of WCA, surface topography, and
abrasion (see Fig. 12b1–b3). Hydrocarbon re-adsorbed on the abraded
presence of adsorbed hydrocarbons for the SPVPS and SPAPS coatings
coating surface in the vacuum treatment process (see Fig. 12b4), and
before and after abrasion are shown schematically in Fig. 12. For the

11
P. Xu, et al. Surface & Coatings Technology 378 (2019) 124929

Fig. 11. Surface chemical compositions of the original coatings, abraded coatings and vacuum-treated coatings: (a) atomic percentage; (b) atomic percentage ratio.

the at.% of hydrocarbon increased to 36.0%, close to that of the original on the coating which was reduced by mechanical contact. Therefore,
SPAPS coating (~38.6%). Different than the flat abraded SPVPS the degraded hydrophobicity of the coatings due to mechanical contact
coating, the residual clusters of the SPAPS coating after abrasion were can be improved after vacuum treatment, although the extent of the
very rough and retained gaps between the residual clusters. Numerous recovery was dependent on the surface structures of the coatings after
fine particles, splats and pores were observed on the fracture surface of mechanical contact. For coatings with rough surface structures, such as
the clusters (see Fig. 8a), creating a hierarchical structure similar to the the SPAPS coating after erosion or linear abrasion, vacuum treatment
original coating. Air could be trapped in the gaps between the residual restored the superhydrophobicity. However, for coatings with relatively
clusters and in the very fine pores of the fracture surface. The Cassie- flat surface topography, such as the SPVPS coating after linear abrasion,
Baxter state was thus re-generated, resulting in a very high WCA and a vacuum treatment increased the WCA of the coating but could not re-
very small RA after the vacuum treatment (see Fig. 12b4). store its original hydrophobicity.
Vacuum treatment can restore the amount of hydrocarbon adsorbed Although the superhydrophobicity, in terms of WCA and RA, was

12
P. Xu, et al. Surface & Coatings Technology 378 (2019) 124929

Fig. 12. Schematic of the variations of WCA and hydrocarbon amount: (a) SPVPS coating; (b) SPAPS coating.

restored for the eroded or abraded SPAPS coating via the vacuum coatings showed small bump surface structures and relatively flat to-
treatment process, the water droplet behavior in the dynamic water pographies in contrast to the porous column-structured coatings de-
impact test changed. As shown in Fig. 13, the impacting water droplet posited via the SPAPS process. The SPVPS coating presented a WCA of
could not rebound or completely rebound from the eroded and abraded ~155° and a much larger RA of ~28° as compared to the super-
SPAPS coating respectively, different than the multiple complete dro- hydrophobic SPAPS coating (WCA of ~163°, RA of ~3°). The SPVPS
plet rebounding behaviors for the original SPAPS coating. This in- coatings maintained high microstructure integrity and relatively con-
dicated that the superhydrophobicity of the abraded coating was de- sistent hydrophobicity after a comprehensive set of mechanical tests,
graded compared to the original coating, although that degradation was showing a much higher mechanical robustness.
not evident in the WCA and RA. The change in wetting behavior of the coatings after mechanical
contacts depends on both the change in surface topography and the
change in surface composition. The investigations of the surface che-
4. Conclusions
mical compositions of the original coatings, the abraded coatings and
the vacuum-treated coatings showed that abrasion reduced the amount
In this work, highly-hydrophobic, mechanically-robust and dense
of the adsorbed hydrocarbon via the removal of coating material (7.4%
Yb2O3 coatings were fabricated by a SPVPS process. The SPVPS

13
P. Xu, et al. Surface & Coatings Technology 378 (2019) 124929

Fig. 13. Dynamic water impact on the eroded and abraded SPAPS coatings after vacuum treatment.

decrease for SPVPS coating after 500 cycles of abrasion; 13.1% decrease 546–551.
for SPAPS coating after 25 cycles of abrasion), while hydrocarbon re- [5] X. Zhu, Z. Zhang, G. Ren, J. Yang, K. Wang, X. Xu, X. Men, X. Zhou, A novel su-
perhydrophobic bulk material, J. Mater. Chem. 22 (2012) 20146–20148.
adsorbed on the surface in the vacuum treatment process, reaching a [6] C. Peng, Z. Chen, M.K. Tiwari, All-organic superhydrophobic coatings with me-
similar level as for the original coatings. The degraded hydrophobicity chanochemical robustness and liquid impalement resistance, Nat. Mater. 17 (2018)
of the abraded coating increased after vacuum treatment, while the 355–361.
[7] X. Chen, Y. Gong, X. Suo, J. Huang, Y. Liu, H. Li, Construction of mechanically
increment was significantly affected by the surface structures of the durable superhydrophobic surfaces by thermal spray deposition and further surface
coatings. The abraded SPAPS coating became superhydrophobic after modification, Appl. Surf. Sci. 356 (2015) 639–644.
vacuum treatment as the hydrocarbon re-adsorbed on the very rough [8] Z. Liu, H. Wang, X. Zhang, C. Lv, C. Wang, Y. Zhu, Robust and chemically stable
superhydrophobic composite ceramic coating repellent even to hot water, Adv.
residual clusters. In contrast, the WCA of the abraded SPVPS coating
Mater. Interfaces 4 (2017) 1601202.
increased mildly from 107° to 126° after vacuum treatment due to the [9] E. Huovinen, L. Takkunen, T. Korpela, M. Suvanto, T.T. Pakkanen, T.A. Pakkanen,
flat coating topography after abrasion. Mechanically robust superhydrophobic polymer surfaces based on protective mi-
cropillars, Langmuir 30 (2014) 1435–1443.
[10] L. Yin, J. Yang, Y. Tang, L. Chen, C. Liu, H. Tang, C. Li, Mechanical durability of
Declaration of competing interest superhydrophobic and oleophobic copper meshes, Appl. Surf. Sci. 316 (2014)
259–263.
None. [11] X. Luo, C. Li, Bioinspired mechanically robust metal-based water repellent surface
enabled by scalable construction of a flexible coral-reef-like architecture, Small
(2019) 1901919.
Acknowledgements [12] M. Long, S. Peng, W. Deng, X. Yang, K. Miao, N. Wen, X. Miao, W. Deng, Robust and
thermal-healing superhydrophobic surfaces by spin-coating of poly-
dimethylsiloxane, J. Colloid Interface Sci. 508 (2017) 18–27.
The work is supported by the China Scholarship Council, Ontario [13] N. Bai, Q. Li, H. Dong, C. Tan, P. Cai, L. Xu, A versatile approach for preparing self-
Research Foundation, NSERC Discovery Grants Program [grant number recovering superhydrophobic coatings, J. Mater. Chem. 293 (2016) 75–81.
RGPIN-2015-06377 (TWC)] and NSERC Green Surface Engineering for [14] G. Azimi, R. Dhiman, H.-M. Kwon, A.T. Paxson, K.K. Varanasi, Hydrophobicity of
rare-earth oxide ceramics, Nat. Mater. 12 (2013) 315–320.
Advanced Manufacturing (Green-SEAM) Strategic Network. [15] D.J. Preston, N. Miljkovic, J. Sack, R. Enright, J. Queeney, E.N. Wang, Effect of
hydrocarbon adsorption on the wettability of rare earth oxide ceramics, Appl. Phys.
Appendix A. Supplementary data Lett. 105 (2014) 011601.
[16] J. Tam, B. Feng, Y. Ikuhara, H. Ohta, U. Erb, Crystallographic orientation-surface
energy-wetting property relationships of rare earth oxides, J. Mater. Chem. A 6
See supplementary data for more details about experimental pro- (2018) 18384–18388.
cedure, coating microstructures and surface chemical compositions. [17] S. Prakash, S. Ghosh, A. Patra, M. Annamalai, M.R. Motapothula, S. Sarkar, S.J. Tan,
J. Zhunan, K.P. Loh, T. Venkatesan, Intrinsic hydrophilic nature of epitaxial thin-
Supplementary data to this article can be found online at https://doi.
film of rare-earth oxide grown by pulsed laser deposition, Nanoscale 10 (2018)
org/10.1016/j.surfcoat.2019.124929. 3356–3361.
[18] E. Külah, L. Marot, R. Steiner, A. Romanyuk, T.A. Jung, A. Wäckerlin, E. Meyer,
References Surface chemistry of rare-earth oxide surfaces at ambient conditions: reactions with
water and hydrocarbons, Sci. Rep. 7 (2017) 43369.
[19] S. Khan, G. Azimi, B. Yildiz, K.K. Varanasi, Role of surface oxygen-to-metal ratio on
[1] M. Ma, R.M. Hill, Superhydrophobic surfaces, Curr. Opin. Colloid Interface Sci. 11 the wettability of rare-earth oxides, Appl. Phys. Lett. 106 (2015) 061601.
(2006) 193–202. [20] M. Bai, H. Kazi, X. Zhang, J. Liu, T. Hussain, Robust hydrophobic surfaces from
[2] E. Celia, T. Darmanin, E.T. de Givenchy, S. Amigoni, F. Guittard, Recent advances in suspension HVOF thermal sprayed rare-earth oxide ceramics coatings, Sci. Rep. 8
designing superhydrophobic surfaces, J. Colloid Interface Sci. 402 (2013) 1–18. (2018) 6973.
[3] T. Verho, C. Bower, P. Andrew, S. Franssila, O. Ikkala, R.H. Ras, Mechanically [21] S. Zenkin, Š. Kos, J. Musil, Hydrophobicity of thin films of compounds of low-
durable superhydrophobic surfaces, Adv. Mater. 23 (2011) 673–678. electronegativity metals, J. Am. Ceram. Soc. 97 (2014) 2713–2717.
[4] A. Cassie, S. Baxter, Wettability of porous surfaces, Trans. Faraday Soc. 40 (1944) [22] P. Xu, T.W. Coyle, L. Pershin, J. Mostaghimi, Fabrication of micro-/nano-structured

14
P. Xu, et al. Surface & Coatings Technology 378 (2019) 124929

superhydrophobic ceramic coating with reversible wettability via a novel solution (2011) 117–132.
precursor vacuum plasma spray process, Mater. Des. 160 (2018) 974–984. [33] P.L. Fauchais, J.V.R. Heberlein, M.I. Boulos, Thermal Spray Fundamentals: From
[23] F. Pedraza, S. Mahadik, B. Bouchaud, Synthesis of ceria based superhydrophobic Powder to Part, Springer Science & Business Media, 2014.
coating on Ni20Cr substrate via cathodic electrodeposition, Phys. Chem. Chem. [34] A. Milionis, E. Loth, I.S. Bayer, Recent advances in the mechanical durability of
Phys. 17 (2015) 31750–31757. superhydrophobic materials, Adv. Colloid Interf. Sci. 229 (2016) 57–79.
[24] J. Tam, G. Palumbo, U. Erb, G. Azimi, Robust hydrophobic rare earth oxide com- [35] X. Tian, T. Verho, R.H. Ras, Moving superhydrophobic surfaces toward real-world
posite electrodeposits, Adv. Mater. Interfaces 4 (2017) 1700850. applications, Science 352 (2016) 142–143.
[25] K. Nakayama, T. Hiraga, C. Zhu, E. Tsuji, Y. Aoki, H. Habazaki, Facile preparation [36] L.R. Scarratt, U. Steiner, C. Neto, A review on the mechanical and thermodynamic
of self-healing superhydrophobic CeO2 surface by electrochemical processes, Appl. robustness of superhydrophobic surfaces, Adv. Colloid Interf. Sci. 246 (2017)
Surf. Sci. 423 (2017) 968–976. 133–152.
[26] Y.J. Cho, H. Jang, K.S. Lee, D.R. Kim, Direct growth of cerium oxide nanorods on [37] A. Steele, I. Bayer, E. Loth, Adhesion strength and superhydrophobicity of poly-
diverse substrates for superhydrophobicity and corrosion resistance, Appl. Surf. Sci. urethane/organoclay nanocomposite coatings, J. Appl. Polym. Sci. 125 (2012)
340 (2015) 96–101. E445–E452.
[27] T. An, X. Deng, S. Liu, S. Wang, J. Ju, C. Dou, Growth and roughness dependent [38] X. Deng, L. Mammen, Y. Zhao, P. Lellig, K. Müllen, C. Li, H.J. Butt, D. Vollmer,
wetting properties of CeO2 films prepared by glancing angle deposition, Ceram. Int. Transparent, thermally stable and mechanically robust superhydrophobic surfaces
44 (2018) 9742–9745. made from porous silica capsules, Adv. Mater. 23 (2011) 2962–2965.
[28] L. Hu, X. Song, D. Jin, C. Xing, X. Shan, X. Zhao, F. Guo, P. Xiao, A robust quasi- [39] X. Deng, L. Mammen, H.-J. Butt, D. Vollmer, Candle soot as a template for a
superhydrophobic ceria coating prepared using air-plasma spraying, J. Am. Ceram. transparent robust superamphiphobic coating, Science 335 (2012) 67–70.
Soc. 102 (2019) 1386–1393. [40] D.S. Kim, A. Suh, S. Yang, D.K. Yoon, Grooving of nanoparticles using sublimable
[29] P. Xu, L. Pershin, J. Mostaghimi, T.W. Coyle, Efficient one-step fabrication of liquid crystal for transparent omniphobic surface, J. Colloid Interface Sci. 513
ceramic superhydrophobic coatings by solution precursor plasma spray, Mater. Lett. (2018) 585–591.
211 (2018) 24–27. [41] N.A. Patankar, Mimicking the lotus effect: influence of double roughness structures
[30] P. Xu, T.W. Coyle, L. Pershin, J. Mostaghimi, Superhydrophobic ceramic coating: and slender pillars, Langmuir 20 (2004) 8209–8213.
fabrication by solution precursor plasma spray and investigation of wetting beha- [42] L. Feng, Y. Zhang, J. Xi, Y. Zhu, N. Wang, F. Xia, L. Jiang, Petal effect: a super-
vior, J. Colloid Interface Sci. 523 (2018) 35–44. hydrophobic state with high adhesive force, Langmuir 24 (2008) 4114–4119.
[31] P. Xu, T.W. Coyle, L. Pershin, J. Mostaghimi, Fabrication of superhydrophobic [43] R.N. Wenzel, Resistance of solid surfaces to wetting by water, Ind. Eng. Chem. 28
ceramic coatings via solution precursor plasma spray under atmospheric and low- (1936) 988–994.
pressure conditions, J. Therm. Spray Technol. 28 (2019) 242–254. [44] J. Long, M. Zhong, H. Zhang, P. Fan, Superhydrophilicity to superhydrophobicity
[32] M.F. Smith, A.C. Hall, J.D. Fleetwood, P. Meyer, Very low pressure plasma spray-a transition of picosecond laser microstructured aluminum in ambient air, J. Colloid
review of an emerging technology in the thermal spray community, Coatings 1 Interface Sci. 441 (2015) 1–9.

15