Bio inspired self-cleaning ultrahydrophobic aluminium surface by laser processing

Abstract

Micro channels and pillars were fabricated by a nanosecond laser source on thin aluminum foil of 100 μm thickness. The wettability of the laser processed μ-patterns were evaluated by the static water contact angle and found to be at the Cassie–Baxter state. The roll-off and static contact angle are 5° and 180° respectively for the μ-pillar structures. The μ-pillar structure which has a well-formed μ-cell structure demonstrated water droplet rebounding and a self-cleaning effect with edible sugar on the laser patterned surface. The elemental analysis suggests that the pulse width, frequency and power density could greatly influence the nature of the new aluminum oxide surfaces formed after laser processing, particularly in relation to the activation of those surfaces against hydroxylation and further chemisorption of organic molecules from air that transforms the surface free energy leading to a transition from a hydrophilic to ultrahydrophobic surface in a short interval of time.

---

1. Introduction

Inspired by the “lotus effect” artificial superhydrophobic surfaces have been fabricated¹ and tested they demonstrate self-cleaning properties.² The lotus effect corresponds to the removal of contamination on the surface by rolling or bouncing water droplets.² Superhydrophobic properties are a key factor, which promotes the self-cleaning ability with a large static contact angle and small hysteresis. A plethora of techniques^3,4 have been adopted for the fabrication of superhydrophobic surfaces on metals⁵ and nonmetals.⁶ The superhydrophobic surfaces are fabricated by introducing geometrical and chemical modification on the surface. In recent years, techniques such as lithography, laser direct writing⁷ processes and electrochemical methods have been effectively implemented for micropatterning metallic surfaces to manipulate the wetting properties.^1,3,8–10 Micropatterns with nanoscale protrusions on the metal surface can transform the wetting properties of the metals from hydrophilic to superhydrophobic.¹¹

In general, two techniques are widely applied to generate the micro patterns, such as top-down¹ and bottom-up¹² approaches. Bottom-up approaches has been identified with rectilinear geometry, whereas top-down technique has curvilinear features.¹³ Both these approaches have multiple steps before the superhydrophobic property is achieved on the metal surface. The primary step is to produce a periodic surface structure followed by surface activation with low energy materials,¹¹ this process is widely accepted as multiple step process. In single step process, the superhydrophobic surface is reached by creating periodic surface structure in one step process, e.g. laser direct writing or laser ablation. Aluminum and its alloys are widely applied as structural engineering material due to its promising physical properties. The wetting property of aluminum alloy has been improved by anodization technique in single step process.³ Chemical technique is also been employed for the transformation of the wetting property of commercially pure aluminum.¹⁴ J. Long et al. investigated the improvement of wetting properties on aluminum by ultra-short laser induced micropattern with respect to time.¹⁵

Micro and nanopatterned metal and ceramic surfaces, primarily hydrophilic has been transformed to hydrophobic or superhydrophobic after an aging process.^1,16 Several mechanisms have been proposed in literature for the transformation wetting property with respect to time, like partial deoxidation of CuO¹⁷ and decomposition of CO₂ into carbon.¹⁸ The transformation on the wetting character on the micro manipulated surface has been measured by the static contact angle measurement. The self-cleaning or the particle removal mechanism has been strongly governed by the degree of water repellency of the solid surface. Depending on the wetting property of the surface, particles or the contamination has the tendency to attach or detach from the water droplets. The energy required to detach a spherical particle on a plain or non-textured interface is directly proportional to (1 − |cos θ₁|)², θ₁is static contact angle¹⁹ of water on the particle.²⁰ Static contact angle¹⁹ on the laser manipulated surface is calculated by the forces between liquid, gas and solid phases at the point of contact line. These forces are governed by surface roughness and surface tension of the material.^21,22 In spite of the available information on the various wetting transformation techniques in literature, the availability of nanosecond (ns) laser patterned ultrahydrophobic metal surface is very limited.

Although, nanosecond (ns) regime laser pulses can create imperfection in micromachining like large heat-affected zones compared to ultra-short laser pulses, dross and recast formation, these laser sources are widely accepted in industry due to its well proven industrial robustness, high pulse energies and frequency, which offers accelerated machining process. The ns laser processed dual geometries on metals at ultraviolet²³ wavelength is one of the less explored feature with respect to improvement of wetting properties, bouncing characteristic and self-cleaning property. In this study, ns laser pulses were applied on flat substrates of aluminum (Al) to fabricate micro patterns. Two different geometries such as μ-channels and μ-pillars have been studied with respect to the SCA. The micro structures of the laser processed samples were characterized by scanning electron microscope (SEM) and confocal laser scanning microscope (CLSM). X-ray photoelectron spectroscopy (XPS) technique has been employed for the quantitative analysis of elemental concentration on the laser patterned surface. Also, the laser processed μ-patterns were subjected to self-cleaning tests.

2. Experimental

Thin sheets of aluminum about 100 μm thicknesses were micromachined by ultraviolet laser pulses. Nano-second laser pulses with primary wavelength of 1064 nm was employed for the laser patterning experiment. The fundamental frequency is frequency-tripled (wavelength of 355 nm) by process of nonlinear optical conversion. The laser source has an average maximum power of 20 W at 100 kHz. The laser beam was guided on to the sample surface by an optical system that has six mirrors, a beam expander and finally to with a lens of 250 mm focal length. The experiments were performed at fixed pulse duration of about 30 ns. The laser beam applied for the experiments has Gaussian power density distribution and the micromachining was executed in atmospheric conditions with a laser spot diameter of 15 μm. Two sets of samples were produced at 500 mW with the geometry of μ-channels and μ-pillars. The microchannels and micropillars are separated by distance in a range of 10–25 μm. The distance is hereafter termed as pitch “P” for future reference in the text. The detailed laser processing conditions are listed in Table 1.

Table 1 Process parameters

Laser power, PL (mW)	500
Spot diameter, D (μm)	15
Fluence, F (J cm^−2)	2.8
Repetition rate, R (kHz)	100
Scan speed, V (mm s^−1)	40
Pitch, P (μm)	10, 15, 20, 25

---

The laser machined μ-patterns were analyzed with SEM attached with energy dispersive X-ray spectroscope (HITACHI, Model: S-3000N) and CLSM to evaluate the geometry of the μ-patterns. The hydrophobicity/water repellence of the samples was investigated by measuring the SCA by sessile drop technique, a video-based optical contact angle measuring device (OCA 15 plus, Data Physics Instruments). Distilled deionized water of 8 μL volume was dispensed on the laser-machined surface structures in atmospheric conditions, and the static contact angle was computed by analyzing the droplet images recorded just after the drop deposition on the laser patterned surface. The change in the surface chemistry on the thin layer of the laser patterned surface has been analyzed by XPS.

3. Results and discussion

3.1. Microstructure

μ-channels and μ-pillars have been fabricated with process parameters shown in Table 1. Fig. 1 represents the CLSM 3D image of the one directional (1D) μ-channels fabricated with four P: 10, 15, 20 and 25 μm. The width of the channel measurements are about of 3, 5, 7 and 8 μm corresponding to the P value of 10, 15, 20 and 25 μm. All the μ-channels recorded a depth in the range of 1.75–2.75 μm. The significant amount of molten material is redeposited around the edges of the μ-channels fabricated. The melt and recast formation is unavoidable in ns laser processing. The depth of the micro channels should be constant irrespective of the pitch distance. However, there has been a difference, may be related to recast formation of the molten materials surrounding the μ-channels. Moreover, the difference in the depth could also be related to the height of the piled recast layer or μ-wall like structure above the substrate surface. The height and width of the μ-wall is influenced by the overlap of the recast layer formed on the adjacent μ-channel edges. The μ-channels fabricated with P: 10 and 15 μm resulted into near broken state. The height of the μ-wall is influenced by melt ejection and the repeated melting of the recast by the low power density tail end of the Gaussian beam profile.

Fig. 1 CLSM 3D image for the microchannels fabricated with, power: 500 mW, scan speed: 40 mm s⁻¹ and repetition rate: 100 kHz for four different P; (A) 10 μm; (B) 15 μm; (C) 20 μm; (D) 25 μm.

From the CLSM image, it is apparent that, the surface has high density of debris on the surface, which is resulted due to splashing of the molten material and redeposition from the vaporized material. Fig. 2 represents the SEM images of μ-channels fabricated with the power of 500 mW. The formation of the recast layer is all along the direction of laser machining. For the sample machined with P: 10 μm, the small spatial shift in the laser beam has produced very narrow μ-channel, with the effective average channel width of 1.75 μm. The narrow width and reduced height can be attributed repeated melting of the recast layer by the tail end of the Gaussian power density distribution. The sample machined with P: 15, 20 and 25 μm has virgin metal surface between the microchannels and it has relatively higher depth (2–2.75 μm). The recast materials from the adjacent μ-channels form an additional structure above the surface similar to wall with height in μm range. The formation of μ-wall forms additional channel like structures above the base material with depth in μm with respect to the surface plain. Therefore, the improvement in the depth of the μ-channels is attributed to the increase in the height of the μ-walls generated by the recast layer.

Fig. 2 SEM image for the μ-channels fabricated with, power: 500 mW, scan speed: 40 mm s⁻¹ and repetition rate: 100 kHz for four different P: (A) 10 μm; (B) 15 μm; (C) 20 μm; (D) 25 μm.

Unlike the sample treated with P: 10 μm, the presence of spherical shaped particle on the top of the recast layer has been observed. The spherical shaped particle is a result of resolidification of metal vapor caused due to the ablation mechanism. Further, the laser patterned region has debris spread over the unprocessed aluminum surface. The dual μ-channels (one below the surface plain and one above the surface plain) are advantageous with respect to improvement of hydrophobicity of the Al surface.

Fig. 3 represents the SEM image of the μ-pillars fabricated with four P: 10, 15, 20 and 25 μm at 500 mW laser power by 2D scanning of the laser beam. The μ-pillars fabricated with P: 10 and 15 μm resulted into scattered or partially scattered structure instead of generating array μ-pillars in regular intervals. There has been heavy deposition of recast material on top surface of the μ-pillars. Whereas, the μ-pillars fabricated with P: 20 and 25 μm have recast deposition over the circumference of the μ-pillars. It is clear that, the molten metal generated by train of laser pulses has been ejected and solidifies along the direction laser movement. For all the four P, the μ-channels are opened in one direction and closed on the other direction. This is caused by the ejection and solidification of the melt formed during the secondary laser scanning direction. In addition to the heavy piling up of recast material on the top edges of μ-pillars, spherical shaped resolidified metal vapor structures are also found. This could be the result of self-cooled metal vapor from the plasma generated by laser ablation. The piling of the recast forms a closed packet “μ-cell” on the top of the micro pillars as shown in Fig. 4(B). The μ-cell structure is separated by μ-channels. Therefore, the laser patterned aluminum surface has dual μ-patterns such as μ-pillars and μ-cell. Thus, reducing the fraction of solid area in contact to the water droplets. The closed μ-channels act as micro packets below the μ-cell structure and it can hold small volume of air trapped inside the pockets. This would offer a high degree of resistance to flush out the small volume of air inside the narrow μ-channels by the water droplets.

Fig. 3 SEM image for the μ-pillar fabricated with, power: 500 mW, scan speed: 40 mm s⁻¹ and repetition rate: 100 kHz for four different P; (A) 10 μm; (B) 15 μm; (C) 20 μm; (D) 25 μm.

Fig. 4 (A) SEM image of hydrophilic region (α) in between the μ-channels, (B) μ-cell formation on top of the micro pillars (P: 25 μm).

3.2. Wetting properties

The effect of μ-patterns on the wetting properties of the Al surface was evaluated by static sessile drop contact angle measurements, with a droplet size of 8 μL. The SCA values with respect to P are presented in Fig. 6. The non-patterned Al surface recorded a SCA value of 85 ± 3°. The laser patterned surfaces were highly hydrophilic immediately after laser micromachining. The SCA measurements were unable to be computed due to very low CA values (<30°). However, the measurements performed after 24 hours exhibited the hydrophobic character. Earlier reports on electrodeposited surface exhibited superhydrophobicity after 2 weeks of exposure to air.²⁴ The SCA measurements on μ-channel patterns shows increase in SCA values with respect to P and it saturates as the P is increased beyond 20 μm. The relatively low values of SCA for the sample processed with P: 10 μm is attributed to the non-formation of periodic μ-patterns. The measurement shows an increase in SCA values with respect to P and starts declination beyond P: 20 μm for μ-pillar structures. When 8 μL volume of deionized water droplet was gently placed on the patterned surface with micro syringe, the droplets does not fall on the surface even after enabling the stroke function in the measuring device. Therefore, an attempt was done to deform the water droplet by the upward push of sample towards droplet. As the laser patterned surface was taken back, the water droplets move away with micro syringe. For the cases where the water droplets are unable to dispense or land on the surface from the micro syringe, the normal sessile droplet technique is not applicable. The SCA of these kind of surface is considered to be 180° (ref. 25) or ultrahydrophobic surface (see ESI Video 1†). This kind of water repellence was observed until the droplet size was increased up to 12 μL. For 12 μL droplet size, the detached water droplet rolls out of the surface as soon as it lands on the patterned surface. The ultrahydrophobicity of the laser patterned surface can be attributed to the dual structure such as μ-pillar and μ-cell structures (Fig. 4(B)) formed on the top of the μ-pillar.

Fig. 5 Images of droplet rolling on ultrahydrophobic Al surface tilted at 5° at different time interval.

Fig. 6 Static contact angle for the micro channels and micro pillars as a function of P.

Samples with dual geometry (μ-pillar and μ-cell) exhibited ultrahydrophobicity for the samples processed with P: 15 and 20 μm. The water droplet from the μ-syringe was made to contact and squished at the laser patterned surface. The water droplet neither wet the surface nor detach from the μ-syringe. The same phenomenon was observed for repeated attempt at the same point. The non-detachment of the water droplet can attributed to the very low value of adhesive force between the droplet and the laser patterned surface (ESI Video 1†). The sample processed at P: 25 μm recorded a SCA of 160°. As the droplet is dispensed on the laser patterned surface, the water droplet has contact only with the top surface of the μ-walls due to the small volume of air trapped in the μ-cell structures. The air bubbles act as cushion for the water droplet and it reduce the probability of direct solid liquid interface which in turn promotes the composite interface of solid, liquid and air as explained in Cassie–Baxter model.²⁶

Thus avoids the direct contact with the bottom surface of the μ-cell structures which is hydrophilic (α – Fig. 4) in nature. However, for P > 25 μm, the surface area of the μ-cell structure is relatively larger. Therefore, the dispensed water droplet has been partially exposed to relatively larger hydrophilic surface (which is marked as (α) on the top of the μ-pillar, Fig. 4(B)). Further, the results suggest that as the water droplet comes in contact with the laser patterned surface it wets hydrophilic regions and dewets the ultrahydrophobic μ-walls. In general, the laser processed surface structures are found to be composed of curvilinear features as a result of Gaussian power density distribution. The formation of μ-cell structure suppresses the curvilinear features and it has become more rectilinear (Fig. 4). The magnitude of capillary forces action around μ-cell features must be directed upward and it is greater than the downward forces resulting from gravity, inertia and Laplace curvature, giving a Cassie state of wetting.²⁶ Moreover, the μ-pillar structure exhibited a high degree of water repellence with respect to μ-channel structure. As the water droplets deposited upon the μ-cell structure on the top surface of μ-pillars with sufficiently large capillary force, the droplets tends to suspend due to the composite interface with solid–liquid and air. Further, due to the high density of the secondary features such as recast layer and the bubble shaped resolidifed metal vapor over the recast layer are primary contact line between the water droplet and the solid surface. The sub-micron features pin down the droplets and prevent the intrusion of the droplet into the hydrophilic region. Therefore, as the water droplets rolls down the laser patterned region, the contact line shifts from one sub-micron structure to the subsequent structures. Whereas, for the μ-channel geometry, due to the curvilinear features in between μ-wall, the advancing contact line penetrates or slides into the interstitial hydrophilic spaces (Fig. 4(A) – (α)) and promotes the decrease in SCA values. Compared to the μ-pillar geometry, the density of the sub-micron features are loosely packed in μ-channel due to the 1D laser scanning. In addition to this, ultrahydrophobic behavior can also be related to high packing factor of the μ-scale features in μ-pillar patterned surface. Except the μ-channel sample processed with P: 10 and 15 μm, all laser patterned samples exhibited a Δθ < 5° and sliding contact angle 3–5°.

The high degree of water repellency is due to the trapping of small volume of air in the asperities leading to the suspension of water drops on the rough surface. The apparent contact angle of this surface can be described by Cassie–Baxter equation.²⁶

cos θ^c = f₁ cos θ₁ + f₂ cos θ₂ (1)

where θ^c is apparent contact angle f₁ and f₂ surface fraction of phase 1 and phase 2 respectively. For the rough surface containing only one type of asperities, and the f is solid surface, the air fraction is given as (1 − f) with θ = 180° for air, the CA angle can be calculated by the following relation.

cos θ^c = f(1 + cos θ) − 1 (2)

Sequential images of droplet rolling on a ultrahydrophobic μ-pillar structures processed with P: 20 μm is presented in Fig. 5. The tilt angle of the surface is about 5°. However, the tilt angle for few samples are below 3°. The high water repellent Al surface exhibited consistency in the tilt angle measurement for repeated measurements and an average of 5 reading has been reported.

3.3. Self-cleaning

Researchers have found that, the lotus leaf has hierarchical surface features with micro-scale in range 10–50 μm and finer features in range of 200 nm to 2 μm.^27,28 The presence of hierarchical surface structures along with the hydrophobic epicuticular wax greatly reduces the adhesion of surface contamination or supports the self-cleaning of the lotus leaf. The ns laser patterned μ-pillar structure has similar kind surface features above and below the surface plane as explained in previous section.

Laser patterned μ-pillar has high degree of water repellence which must well support the water bouncing character. The water bouncing study was performed with a high speed camera (Photron) by capturing image at the rate 250 frames per second. The laser patterned μ-pillar structures processed with P: 15 and 20 μm were subjected to evaluate the self-cleaning efficiency or removal of contamination by water droplets as it exhibited ultrahydrophobicity with 8 μL water droplets.

The walls of the μ-cell have protuberance in range of 0.5–1 μm caused by the redeposition of the metal vapour. The protrusions on the micro wall even reduce the effective contact area between the solid and liquid, and promote the solid–air–liquid interface leading to ultrahydrophobic surface. Initially, bouncing behavior of the water droplets was studied. Fig. 7 represents the snapshots at different point of time to show the rebound behaviors of the surface. A regular medical syringe has been used to drop the water droplets normal to the laser patterned surface from a height of 5 cm. The laser patterned surface is placed on the plain surface without any tilt angle. It is important to note that, regular tap water has been used to simulate the real time applications. Therefore, the water may have the chemical contamination like chlorine, minerals etc. The observation of the snapshots clearly shows the bouncing behavior of the water droplets. The water droplets are highly repelled and it bounces one or two times before rolls off the laser patterned region (see ESI Video 2†) or forms bigger droplets by cascading coalescence process.

Fig. 7 Bouncing and coalescence of water droplet striking the ultrahydrophobic aluminum surface.

The bouncing of the water can be attributed to the presence of air between the droplet and the solid surface.²⁹ The rebounding height on laser patterned region has been considerably reduced due to the loss of kinetic energy of the droplet. The mass of the water droplet is another prime factor with determines the number of bounces. As the size of the water droplet is bigger, the depth of penetration inside the μ-patterns would also increases which in turn reduce the number of bounces. As shown in Fig. 7(B), for some of our experiments, we found break up in droplets on the first bounce and the smaller fragmentation remains in contact {Fig. 7(D)} with the surface. However, in few cases, the smaller droplets recombine with larger fragmentation and rolls out of the surface. The high degree of the water repellence can be related dual geometry and trapping or presence of air in the micro voids created by laser patterning. Another factor required for the high degree of water repellence is low surface energy on the laser patterned region. The surface micro patterns with rounded edge (Fig. 4) are considered to be better than the sharp edge μ-pattern as for as the bouncing efficiency²⁹ is concerned.

In lotus leaf, self-cleaning is performed by rolling water droplets. The rolling droplets on superhydrophobic surface is more efficient than the sliding droplets for the removal of contaminations.³⁰ Fig. 8 shows the snapshots on the self-cleaning process. In our experiments, we employed edible sugar for the tests. Sugar particles is sprinkled on the surface and water droplets are made to strike the surface about 90° with respect to the laser structured surface. Primarily, coalescence of water drops after striking the hydrophobic surface (see ESI Video 3†) is observed. The cascading coalescence process forms bigger water droplets and tend to move away from the laser patterned area.

Fig. 8 Snapshots of aggregating removal process.

Coalescence process enables the water droplets to form a liquid bridge among the neighboring particles and transforms into a cluster. Eventually, this cluster moves away from the surface. The cascading coalescence effect has been exploited to self-clean the surface with ultrahydrophobic material surface. The surface remains clean and dry after the removal of the sugar particles. The cleaning process is termed as aggregating removal process (see ESI Video 3†). The cleaning process in our experiments is very significant, as the viscosity and surface tension of the water droplet has been changed due to the mixing of sugar. The self-cleaning experiments has been reported on fs processed platinum samples by applying dust particle.³¹ In the reported work, self-cleaning experiments were performed at a tilt angle 8°, whereas, the present experiments are performed without tilt angle. The experiments have been repeated several times to validate the reproducibility and durability of the self-cleaning process.

3.4. Surface chemical analysis (XPS)

The secondary reason for the transition of the wetting properties could be the factor of change in surface chemistry on the laser processed samples. In previous works, J. Long,¹⁵ Kietzig³² have reported transitions to superhydrophobicity related to changes of surface chemical composition in laser processed metallic substrates after ambient exposure due to the formation of a carbonaceous layer on top of hydrophilic metal oxide surface. In order to investigate this point in our case, XPS analysis of samples before and after laser processing were carried out. The XPS elemental analysis is restricted only to the μ-pillar structures which exhibited ultrahydrophobicity.

Fig. 9 represents spectra for untreated sample and laser patterned μ-pillar structure at P: 15 μm. All the laser patterned samples were stored in ambient air to simulate the industrial environment. The XPS spectrum clearly shows the presence of three major peaks at 531.5, 284.5 and 74 eV, corresponding to O 1s, C 1s, and Al 2p, respectively. All the laser processed samples show these three peaks at the same binding energies. The binding energies of O 1s and Al 2p peaks indicate the presence of Al₂O₃, whereas the C 1s peak is attributed to hydrocarbon contamination of the surface.

Fig. 9 XPS spectrum for Al (A) Unprocessed and (B) Laser processed with μ-pillar geometry.

The elemental analysis of the untreated and laser micromachined Al samples are shown in Table 2. In the case of the unprocessed Al substrate, it is well known that a thin passivation layer of Al₂O₃ is formed on the aluminum surface when it has been exposed to the atmospheric air. Its high carbon at% can be attributed to hydrocarbon contamination by a prolonged exposure to the atmosphere. However, as it can be observed, the ratio of O/Al is much higher than the stoichiometric one for Al₂O₃, indicating that the contamination could also be due to the use of organic solvent containing oxygen, like acetone, as cleaning and degreasing agent of commercial aluminum foils. In the case of laser processed samples the carbon content is very similar but lower than the untreated one, even after three days under atmosphere exposure, which can be related to a cleaning effect of the surface by the laser treatment. The elemental concentration of aluminum and oxygen are also similar for the samples recorded superhydrophobic (P: 10 and 25 μm) and ultrahydrophobic (P: 15, 20 μm) contact angles, but as it can be observed from Table 2, the O/Al ratio is higher than that expected for Al₂O₃ (1.5). This fact could indicate that together Al₂O₃, other aluminum oxides like AlO(OH) can be formed during the laser processing, as reported by Balchev et al.³³ The presence on the surface of hydroxyl groups due to hydroxylation reactions and absorbed water molecules once the samples are exposed to the moisture of the atmosphere can also contribute to the high O/Al ratio observed.

Table 2 The chemical composition of micro pillars fabricated with 500 mW laser power and the corresponding static contact angle

Hatch distance, P micro pillars (μm)	C (at%)	O (at%)	Al (at%)	O/Al	C/Al	SCA (θ)
10	41.42	42.55	14.38	2.96	2.88	148
15	34.54	46.36	17.93	2.58	1.93	180
20	34.29	47.29	17.71	2.67	1.94	180
25	32.86	47.93	17.91	2.67	1.83	160
Un processed surface	75.60	17.72	2.44	7.26	30.98	87

---

It is well established that the affinity of solid surfaces for water molecules, and therefore their hydrophilic character, depends on the surface polarity, so normally, higher the polarity greater would be the hydrophilicity and wettability. The surface of common transition metal oxides, as alumina, has a large number of coordinatively unsaturated aluminium and oxygen atoms (hereafter referred as cus: coordinatively unsaturated sites) that act as Lewis acid and base sites, respectively, which are directly related to the polar surface free energy. This is the reason for the hydrophilicity observed in the majority of ceramics and metals. Therefore, in the case of aluminum oxide, there are many acid–base pairs at the surface that can participate in the heterolytic dissociative adsorption of water molecules from the moisture in the atmosphere to satisfy the coordinatively unsaturated Al³⁺ cations. Aluminum ions, acting as Lewis acids, can accept a lone electronic pair from the oxygen atom of water molecules that dissociate upon adsorption in H⁺ and OH⁻. Based on the reactions sequence proposed by Kung³⁴ for transition metal oxides, the suggested water heterocyclic dissociation onto alumina surface is as follows:

Al³⁺ (cus) + H₂O ⇌ Al³⁺OH₂ (3)

In eqn (3), water is adsorbed molecularly to satisfy the coordinatively unsaturation of the metal ion. Then, eqn (4), the dissociative adsorption of a water molecule as OH⁻ and H⁺ by the coordinatively unsaturated cation–anion pair takes place, with the proton transferred to a nearby surface base site (an oxygen atom). As Hass³⁵ and Wippermann³⁶ have reported, for low water coverage dissociate single water molecules are energetically most favored, however, as the water coverage increases, for instance after a prolonged exposure to the moisture in the atmosphere air, more complex structures composed of dissociate and undissociate water monomers, which form hydrogen bonds with pre-adsorbed H/OH groups, are favored. According to Wippermann,³⁶ adsorption energies suggest that, the molecule–surface interaction is dominant over the molecule–molecule interaction. The hydroxylation of freshly formed aluminum oxide after laser processing, quickly and effectively passivates the polar sites on the aluminum oxide surface, leading to a weakening of the strength of the Lewis acid–base pairs that reduces the total surface free energy of the oxide layer. Therefore, a lower hydrophilicity and wettability with the corresponding increase of the contact angle is observed.^37,38 Hydroxylation of the oxide layer surface seems to occur through a dissolution–precipitation mechanism, leading to the formation of an amorphous aluminum oxyhydroxide region.³⁹ This could also explain the high at% O/Al ratio observed by XPS analysis.

Another way to reduce the surface polarity of aluminum oxide, even in a more efficient manner according to the results obtained by Long,¹⁵ is the adsorption of organic molecules from air because they are essentially nonpolar, leading to the transition in the surface wettability with the increase in the CA. Several authors^15,39–43 have reported about the presence of hydroxyl groups on the oxide surface plays an significant role, due to the chemisorption of small, short chain organic species which are present in ambient air, for instance acetic acid or formic acid,^44,45 primarily takes place through the interactions with hydroxyl groups, yielding chemisorbed carboxylate.^39,43 However, it has been reported that the adsorption of non-dissociated water molecules which are hydrogen bonded to the preadsorbed OH/H sites, hinder the further adsorption of airborne hydrocarbons by blocking the surface active sites.^39,43 Both the hydroxylation and the chemisorption of organic species from the atmosphere can explain the transition from hydrophilicity to superhydrophobicity of the laser processed aluminum substrates after atmospheric exposure for three days. However, in order to explain the transition to superhydrophobicity and ultrahydrophobicity in the case of μ-pillar structured samples, as well as the differences between their contact angles, the laser pattern modifications must be taken into account as a synergistic key factor in combination with the changes in surface chemical composition. The C/Al ratio (see Table 2) can be used as a measure of the relative amount of adsorbed organic molecules, since when hydrocarbons are adsorbed onto the aluminum oxide surface, less aluminum and more carbon are detected because the analysis depth of XPS is small, approximately 7 nm for C 1s and 8 nm for Al 2p photoelectrons.⁴⁶ Therefore, from a qualitative point of view, it can be related to the non polar level of the laser processed surfaces. The P: 10 μm sample shows the highest C/Al ratio, but as it can be observed from Table 2 has the lowest SCA, 148°, of the four μ-pillar structures considered, even smaller than P: 25 μm sample, CA: 160°, which possesses the lowest C/Al ratio, indicating a higher polar surface free energy for the latter. On the other hand, P: 15 and P: 20 μm samples have very similar C/Al ratios, but are approximately 50% lower than P: 10 μm values; however, they show the greatest contact angles, 180°, which are consistent with an ultrahydrophobic behavior. As it can be observed, the wetting response of the μ-pillar structured samples is not that were expected for the different polarity and surface free energy associated with the surface chemical composition. Therefore, it can be concluded that the transition from superhydrophobicity to ultrahydrophobicity should be mainly explained by the differences in the surface texturing; and the changes in the surface chemical composition have a little effect on this particular wetting transition for μ-pillar structures. As it has been mentioned, the μ-pillar structure P: 10 μm has the highest C/Al ratio, so it suggests that it has been able to adsorb organic material in larger amounts than the other μ-pillar structures after exposure to the same atmosphere for the same time (probably because of its surface area is larger); thus it should be the most hydrophobic from a surface chemical composition approach. However, as it was mentioned in Section 3.2, the laser pattern does not show a well-defined dual geometry composed by μ-pillars and μ-cells (see Fig. 3), but an irregular one with a rough surface. This fact could move away the wetting state from the ideal Cassi–Baxter model, leading to a high contact angle, but not ultrahydrophobic. In the case of μ-pillar structures P: 15, 20 and 25 μm, a regular laser pattern are formed with μ-pillars and μ-cells (see Fig. 3). Nevertheless, P: 25 μm sample shows μ-cells that have a relatively larger surface area (Fig. 4), so the water droplets have been partially exposed to a higher hydrophilic surface, which drops its contact angle until 160°, changing its wetting properties from ultrahydrophobic to superhydrophobic.

In the present study, the Al surface shows higher C/Al ratios and needs lower exposure times to air, 2–3 days, to reach the ultrahydrophobic state than those found in a previous work¹⁵ for ps laser patterning Al samples stored in air, where an aging time of 30 days to attain the transformation of the wetting properties from hydrophilic to superhydrophobic is reported. Therefore, the pulse duration, pulse frequency and power density could have a critical influence in the nature of the new aluminum oxide surfaces formed after laser processing, particularly in relation to the activation of those surfaces against hydroxylation and further chemisorption of organic molecules from air processes, that change the surface free energy and thus the hydrophilic properties of the surfaces.

4. Conclusion

μ-channels and μ-pillars were successfully fabricated by ns laser source on thin aluminum foil of 100 μm thickness. The wetting property analysis shows a clear transition from hydrophilic to hydrophobic in 24 hours and ultrahydrophobic within 72 hours from the time of laser micro-fabrication for the μ-pillar geometry with μ-cell structure on the top of the pillars without any additional low surface energy coatings. The highest SCA recorded for the laser processed μ-patterns are 180° for 8 μL droplet volume and the roll-off angle is found to be 5°. The ns laser processing has replaced typical curvilinear structure with rectilinear features. The μ-pillar structure which has well-formed μ-cell structure demonstrated water droplet rebounding and self-cleaning effect with edible sugar on the surface. The XPS elemental analysis suggest that the pulse width, frequency and power density could greatly influence the nature of the new aluminum oxide surfaces formed due to laser ablation, particularly in relation to the activation of those surfaces against hydroxylation, and chemisorption of organic molecules from air. This surface chemical change reduces the surface free energy, leading to a transition from hydrophilic to ultrahydrophobic surface in short interval of time. The wetting property transition by direct laser writing technique in ns regime has potential industrial applications.

Acknowledgements

The authors are thankful to BSH Electrodomésticos España for their support in executing this work.

References

1. R. Jagdheesh, Langmuir, 2014, 30, 12067–12073.
2. R. Fürstner, W. Barthlott, C. Neinhuis and P. Walzel, Langmuir, 2005, 21, 956–961.
3. H. Li, S. Yu, X. Han, E. Liu and Y. Zhao, Colloids Surf., A, 2015, 469, 271–278.
4. S. S. Latthe, C. Terashima, K. Nakata, M. Sakai and A. Fujishima, J. Mater. Chem. A, 2014, 2, 5548–5553.
5. R. Jagdheesh, B. Pathiraj, E. Karatay, G. R. B. E. Römer and A. J. Huis in't Veld, Langmuir, 2011, 27, 8464–8469.
6. J. Lee and R. S. Fearing, Langmuir, 2012, 28, 15372–15377.
7. R. Jagdheesh, J. J. García-Ballesteros and J. L. Ocaña, Appl. Surf. Sci., 2016, 374, 2–11.
8. M. Miwa, A. Nakajima, A. Fujishima, K. Hashimoto and T. Watanabe, Langmuir, 2000, 16, 5754–5760.
9. G. Li, Z. Li, L. Lu, L. Xue and C. Deng, J. Wuhan Univ. Technol., Mater. Sci. Ed., 2012, 27, 138–141.
10. Z. Yuan, J. Xiao, C. Wang, J. Zeng, S. Xing and J. Liu, J. Coat. Technol. Res., 2011, 8, 773–777.
11. R. Jagdheesh, B. Pathiraj, E. Karatay, G. R. B. E. Romer and A. J. Huis in't Veld, Langmuir, 2011, 27, 8464–8469.
12. S. S. Latthe, P. Sudhagar, C. Ravidhas, A. Jennifer Christy, D. David Kirubakaran, R. Venkatesh, A. Devadoss, C. Terashima, K. Nakata and A. Fujishima, CrystEngComm, 2015, 17, 2624–2628.
13. C. W. Extrand and S. I. Moon, Langmuir, 2014, 30, 8791–8797.
14. Y. Liu, J. Liu, S. Li, Y. Wang, Z. Han and L. Ren, Colloids Surf., A, 2015, 466, 125–131.
15. J. Long, M. Zhong, H. Zhang and P. Fan, J. Colloid Interface Sci., 2015, 441, 1–9.
16. A.-M. Kietzig, M. N. Mirvakili, S. Kamal, P. Englezos and S. G. Hatzikiriakos, J. Adhes. Sci. Technol., 2011, 25, 23.
17. F.-M. Chang, S.-L. Cheng, S.-J. Hong, Y.-J. Sheng and H.-K. Tsao, Appl. Phys. Lett., 2010, 96, 114101.
18. A.-M. Kietzig, S. G. Hatzikiriakos and P. Englezos, Langmuir, 2009, 25, 4821–4827.
19. F. Di Niso, C. Gaudiuso, T. Sibillano, F. P. Mezzapesa, A. Ancona and P. M. Lugarà, Opt. Express, 2014, 22, 12200–12210.
20. B. Binks and T. Horozov, Colloidal particles at liquid interfaces, Cambridge Univ press, Cambridge, UK, 2006.
21. A. B. D. Cassie and S. Baxter, Trans. Faraday Soc., 1944, 40, 546–551.
22. K. Liu and L. Jiang, Nanoscale, 2011, 3, 825–838.
23. E. Huovinen, L. Takkunen, T. Korpela, M. Suvanto, T. T. Pakkanen and T. A. Pakkanen, Langmuir, 2014, 30, 1435–1443.
24. S. Khorsand, K. Raeissi, F. Ashrafizadeh and M. A. Arenas, Chem. Eng. J., 2015, 273, 638–646.
25. J. Zhang, X. Gao and L. Jiang, Langmuir, 2007, 23, 3230–3235.
26. A. B. D. Cassie and S. Baxter, Trans. Faraday Soc., 1944, 40, 546–551.
27. K. Koch, B. Bhushan and W. Barthlott, Soft Matter, 2008, 4, 1943–1963.
28. W. Barthlott and C. Neinhuis, Planta, 1997, 202, 8.
29. C. R. Crick and I. P. Parkin, J. Mater. Chem. A, 2013, 1, 799–804.
30. V. Rossbach, P. Patanathabutr and J. Wichitwechkarn, Fibers Polym., 2003, 4, 8–14.
31. A. Y. Vorobyev and C. Guo, J. Appl. Phys., 2015, 117, 033103.
32. A.-M. Kietzig, M. Negar Mirvakili, S. Kamal, P. Englezos and S. G. Hatzikiriakos, J. Adhes. Sci. Technol., 2011, 25, 2789–2809.
33. I. Balchev, N. Minkovski, T. Marinova, M. Shipochka and N. Sabotinov, Mater. Sci. Eng., B, 2006, 135, 108–112.
34. Studies in Surface Science and Catalysis, ed. H. K. Harold, Elsevier, 1989, vol. 45, pp. 53–71.
35. K. C. Hass, W. F. Schneider, A. Curioni and W. Andreoni, Science, 1998, 282, 265–268.
36. S. Wippermann, W. G. Schmidt, P. Thissen and G. Grundmeier, Phys. Status Solidi C, 2010, 7, 137–140.
37. M. M. Gentleman and J. A. Ruud, Langmuir, 2010, 26, 1408–1411.
38. N. Giovambattista, P. G. Debenedetti and P. J. Rossky, J. Phys. Chem. B, 2007, 111, 9581–9587.
39. J. van den Brand, S. Van Gils, P. C. J. Beentjes, H. Terryn and J. H. W. de Wit, Appl. Surf. Sci., 2004, 235, 465–474.
40. M. Psarski, J. Marczak, J. Grobelny and G. Celichowski, J. Nanomater., 2014, 2014, 1–11.
41. H. Knözinger and P. Ratnasamy, Catal. Rev.: Sci. Eng., 1978, 17, 31–70.
42. S. Takeda, M. Fukawa, Y. Hayashi and K. Matsumoto, Thin Solid Films, 1999, 339, 220–224.
43. S. R. Tong, L. Y. Wu, M. F. Ge, W. G. Wang and Z. F. Pu, Atmos. Chem. Phys., 2010, 10, 7561–7574.
44. L. Grönberg, Y. Shen and J. Å. Jönsson, J. Chromatogr. A, 1993, 655, 207–215.
45. E. Dabek-Zlotorzynska and M. McGrath, J. Anal. Chem., 2000, 367, 507–518.
46. M. R. Alexander and T. M. Duc, J. Mater. Chem., 1998, 8, 937–943.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra12236a

---