Superhydrophobic and self-cleaning aluminium surfaces via affordable nanosecond fibre laser and vacuum treatment

Journal of Micromechanics and Microengineering PAPER • OPEN ACCESS Superhydrophobic and self-cleaning aluminium surfaces via affordable nanosecond fibre laser and vacuum treatment To cite this article: K Kobliha et al 2025 J. Micromech. Microeng. 35 105009 View the article online for updates and enhancements. You may also like Design and performance evaluation of a single piezoelectric stacked drive jet dispensing valve Jingwei Yang, Yifan Li, Liangjun Xiang et al. - Measurement of Earth tides with a MEMS gravimeter via releasing the axial force of cross-configured geometric anti-spring Weidong Wang, Taotao Ding, Yiming Chen et al. - Effect of the surface roughness of chamfer on edge-bead effects in spin-coating film formation Ying Yan, Han Li, Qiuyu Liu et al. - This content was downloaded from IP address 147.231.2.11 on 30/10/2025 at 12:21 Journal of Micromechanics and Microengineering J. Micromech. Microeng. 35 (2025) 105009 (10pp) https://doi.org/10.1088/1361-6439/ae12fe Superhydrophobic and self-cleaning aluminium surfaces via affordable nanosecond fibre laser and vacuum treatment K Kobliha1,2,∗, P Hauschwitz2, R Bicistova2, M Prochazka2, J Brajer2and T Mocek2 1Faculty of Nuclear Sciences and Physical Engineering, Czech Technical University in Prague, Brehova 7, 115 19 Prague, Czech Republic 2Hilase Centre, Institute of Physics, Academy of Sciences of the Czech Republic, Za Radnici 828, 252 41 Dolni Brezany, Czech Republic E-mail: [email protected] Received 2 July 2025, revised 8 October 2025 Accepted for publication 14 October 2025 Published 24 October 2025 Abstract Self-cleaning surfaces have a great potential for applications in the construction, aviation, food and energy industries. A common strategy for achieving self-cleaning functionality is to render the surface superhydrophobic, enabling water droplets to roll off and remove dirt particles. In this article the aluminium sample was treated using an affordable nanosecond laser combined with vacuum processing to induce superhydrophobicity. Various processing parameters were investigated, including structural geometry, hatch distance, and additional defocused irradiation. Following fabrication, the surfaces were evaluated for self-cleaning performance by applying MnO2and polyamide particles of defined sizes and tilting the samples by 15◦. A specified number of 20µl water droplets was allowed to roll over the contaminated surfaces and the cleaning efficacy was quantitatively analysed through optical imaging and image processing. The most effective configuration, rhombic structures with 80µm hatch distance, removed over 95% MnO2particles after only three droplets. The confocal and electron microscopy characterisation of the surface was used to understand the relation between surface topography and the cleaning effect. Keywords: self-cleaning surface, laser ablation, microstructuring, superhydrophobic structures, contact angle 1. Introduction The principle of superhydrophobic self-cleaning surfaces is as follows: the water droplet, which is unable to stick to the ∗Author to whom any correspondence should be addressed. Original Content from this work may be used under the terms of the Creative Commons Attribution 4.0 licence. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI. surface, moves down the material and binds impurities from the surface to itself and in the process cleans it. The effectiveness of self-cleaning depends on the properties of the surface, the chemical composition of the contaminants (e.g. whether they are water-soluble), the size of the dirt particles, and the number and size of droplets impacting the surface [1]. The self-cleaning surfaces can be used as the outer layer of the solar panels so that impurities flow from their surface. This ensures their efficient operation even in dusty conditions. They are also suitable for applications in the food industry where they can serve as hygienic worktops in food processing. Potentially, the body of aeroplanes or other objects 1 © 2025 The Author(s). Published by IOP Publishing Ltd J. Micromech. Microeng. 35 (2025) 105009 K Kobliha et al moving through the atmosphere can have self-cleaning properties which significantly facilitate their cleaning and prevent the buildup of dirt on the surface [2]. The superhydrophobic surface is characterised by a large contact angle θ > 150◦, a low sliding angle <10◦, and a low contact angle hysteresis. There are three classic theories describing how a liquid droplet wets a solid surface: Young–Dupré’s theory, Wenzel’s theory, and Cassie–Baxter’s theory [3]. The Young-Dupré model for solids does not take into account any roughness, contamination, or protrusions on the surface [3]. Additionally, it describes the droplet only in a static state and does not consider contact angle hysteresis and therefore is not suitable for the description of laser-treated surfaces. The Wenzel model describes a chemically homogeneous rough surface. According to Wenzel, a droplet of fluid fully penetrates the grooves of the solid. A water droplet attached to the surface exhibits significant hysteresis preventing it from flowing off the surface even at larger angles [4]. The Cassie–Baxter model can additionally describe a chemically heterogeneous surface. The model is based on the idea that a droplet rests on the rough protrusions of the surface making contact with both solid and gas phases [5]. For a homogeneous rough surface (i.e. a surface composed of solid and air components) the water droplet touches only the tips of the protrusions and the air beneath them without fully contacting the solid surface. In this model droplets can easily slide off the surface resulting in low sliding angles and minimal contact angle hysteresis [6]. To produce a superhydrophobic surface it is necessary to reach the Cassie–Baxter state [2]. Many methods have been developed to produce superhydrophobic surfaces including lithography [7], electrodeposition [8], chemical etching [9], thermal embossing [10], chemical vapour deposition [11], and sol– gel methods [12]. However, these methods are procedurally complicated and are not versatile when it comes to the production of structures as they have certain limiting parameters (e.g. long production time, use of chemicals). As an alternative to these methods laser ablation enables the superhydrophobicity of the surface directly without using chemicals [13]. Surface treatment using laser ablation is nowadays widely used in industry due to its speed, relative simplicity, safety, and great flexibility of different types of surfaces [14]. Laser-treated surfaces can have superhydrophobic [1,15–18], antibacterial [15], anti-freezing [16], non-corrosive [17], selfcleaning [1], and many other properties that can be used e.g. in the production of sensors, batteries, optical devices, in the food, aviation, construction, electrotechnical, and energy industries [18]. The laser ablation can be applied to produce structures with a high contact angle and low sliding angle. The laser structures are created through the direct interaction of high-energy laser beams with the material. This process can create structures with micron-level detail although it has certain limitations in terms of processing speed [1,19,20]. To further increase the contact angle of microstructures, the structures can be irradiated with a defocused beam creating protrusions at the nanometer scale on the structured surface. This modified surface exhibits a higher contact angle compared to unmodified surfaces [21]. The superhydrophobicity of surfaces depends not solely on their topography but also on their chemical composition [22]. As a result, the structures are initially hydrophilic after fabrication, and an additional process is necessary to reduce the high surface energy to achieve superhydrophobicity. Hightemperature oxides formed during laser fabrication have high surface energy [23] and can react with atmospheric water molecules (H2O) to produce hydroxyl groups (OH−), creating a passive layer that can further adsorb water via hydrogen bonding. This layer slows down the attachment of airborne hydrocarbons, delaying the transition from hydrophilic to superhydrophobic behaviour. While surface energy can eventually be reduced simply by exposure to air, this process may take weeks to months. A more effective, chemical-free method, as demonstrated by Jagdheesh et al [24], is to place the samples in a vacuum chamber at pressures p<10−4Pa. Under these conditions, the presence of water vapour is significantly reduced, and the absorption of airborne hydrocarbons, originating from the mineral oil in the chamber’s rotary pump, is enhanced by the reduced passive layer on the sample surface [13]. Self-cleaning surfaces via laser structuring were created in the past. Srin et al [25] fabricated a surface with a hierarchical trench pattern and pillar structures on stainless steel using a 100 fs Ti:sapphire laser with a 10 kHz repetition rate. Milles et al [1] created structures on aluminium using nanosecond pulses direct laser writing, picosecond pulses direct laser interference pattering, and a combination of both technologies. Lorenz et al [26] made structures on stainless steel using 12 ps laser and after laser processing treated the surface using phosphonic acid or tetrahydro-furan solution. Yao et al [27] used 35 fs Ti:sapphire laser to manufacture structures on stainless steel. Patil et al [28] fabricated structures on silicone rubber by a nanosecond Nd:YAG laser with a repetition rate of 3 kHz. Despite these promising results, there is still a lack of studies where contact angles above 150◦have been achieved and which deal with the production of superhydrophobic surfaces with a self-cleaning effect. Additionally, in these previously conducted experiments, mostly expensive pico-/femto -second lasers were used and the surface was usually treated with complex chemicals. In this paper superhydrophobic surfaces with self-cleaning functionality are fabricated using an off-the-shelf nanosecond fibre laser (costing less than 10 000 €) in combination with vacuum processing. Moreover, no chemical post-processing is required making the method efficient, cost-effective, and environmentally friendly, while remaining simple to implement and capable of delivering promising functional results. 2. Materials and methods In this paper aluminium alloy 7075 plates were used to fabricate superhydrophobic structures with self-cleaning 2 J. Micromech. Microeng. 35 (2025) 105009 K Kobliha et al properties. Before laser processing all samples were polished to a roughness Ra <0.05µm and cut to approximately 50 ×50 mm2. Samples were treated in an open atmosphere using an off-the-shelf nanosecond ytterbium-based pulsed fibre laser JPT YDFLP-E2-60-M7-M-R emitting 9 ns pulses with a wavelength of 1064 nm. Generated pulses had an energy of approximately 25 µJ with a repetition rate of 100 kHz. The laser beam was directed through a galvanometric scanner Sino-Galvo SG2203 and focused onto the sample using a 210 mm F-theta lens which resulted in a spot diameter of 50 µm. Laser fluence at the focal point was approximately 2.55 J cm−2. The beam was displaced on a sample with a scanning speed of 700 mm·s−1. Immediately after laser processing samples were stored for 12 h in a vacuum chamber pumped down to 5 ·10−6Pa for transfer of initial superhydrophilic surface properties to superhydrophobic [23]. The surface topography of micro and nanostructures was investigated using a confocal microscope Olympus LEXT OLS5000 with 2k×magnification and a scanning electron microscope TESCAN MIRA at an electron energy of 15 keV and working distance ≈10mm with 2k×, 5k×or 15k× magnification. The wettability was evaluated by measuring the static contact angle and sliding angle with an optical contact angle measuring device OCA 15EC. The volume of water droplets was 20 µl which were brought to the surface by the slow motion of the movable rack towards the droplet until the droplet had separated from the dispenser needle. After measuring the contact angle, the device was gradually tilted until the water droplet began to slide. The results were acquired through the average of three measurements on different locations for every sample surface. To determine the self-cleaning properties the surface was contaminated with two different impurities, namely manganese dioxide MnO2with a particle size of 10 µm (SigmaAldrich Ltd) and polyamide with a particle size of 55 µm (LaVision GmbH). These impurities were measured separately on the surface to determine the self-cleaning effect for different particle sizes and chemical compositions. These types of impurities were selected as representative due to their composition (MnO2as an inorganic compound, polyamide as an organic), defined particle sizes (for a more precise determination of self-cleaning), and relatively good availability. Impurities were applied to the surface using a 3D-printed plate with a slit of size 4 ×1 mm2and a height of 0.3 mm (figure 1(a)) using a measuring cup and a spatula (figure 1(b)). A spot of a defined size with approximately the same amount of particles was formed on the surface of the structures each time (figure 1(c)). Such a sample was then placed on the apparatus to measure the self-cleaning effect (figure 1(d)). The self-cleaning effect was measured on an apparatus with the OCA 15EC contact angle measuring device, which was tilted at an angle of 15◦(figure 2(a)). A sample with impurities was placed on the movable table and illuminated using a light source. The specimen was captured using a Basler ACE acA4024-29um camera on which it was attached a 40 mm intermediate camera ring and BASLER C23-5028-5 M lens with a focal length of 50 mm. The resolution of the camera was 2048 ×1088 px2. Impurities were removed from the surface by a droplet of 20 µl which spontaneously flowed on the sample from the dispenser needle. Impurity condition on the sample was captured always at the beginning and after each droplet, while after each a total of 5 droplets flowed onto the sample (or 8 droplets for square structures). The images themselves were processed in ImageJ for each surface and impurity type. 3. Results and discussion 3.1. Laser processing of samples Aluminium samples were treated with laser ablation to produce diverse structures of trenches and hills on the surface. The various surface structure patterns were analysed to assess how structural geometry affects the self-cleaning properties. Selected structures were additionally irradiated with a defocused laser beam after laser ablation to create nanometric protrusions. These protrusions increase surface roughness which leads to an increase in the contact angle (according to the Cassie–Baxter model). Based on the literature and preliminary experiments the initial parameters for laser ablation were selected. The laser was displaced on the surface with the scanning speed of 700 mm·s−1in various directions creating patterns such as squares, honeycombs, circles, triangles, rhombuses, and rectangles (figures 2(b)–(g)). Each pattern was written with 80 overscans meaning the laser beam passed along the same path 80 times. The hatch distance Λvaried between 35 µm and 150 µm resulting in structures of different sizes and geometries. The topography of square structures for different hatch distances is shown in figures 3(a)–(c) while figures 3(d)– (i) presents the topographies of different patterns for hatch distance 100 µm. In nanosecond laser processing material removal occurs mainly due to thermal mechanisms including localised melting, molten material flow, and subsequent resolidification. As the molten material redistributes it tends to accumulate along the edges of the laser path forming elevated overhangs around the ablated regions (e.g. figures 3(a)–(c)). The details of this redistribution, and the resulting hierarchical microand nanostructures, are strongly influenced by laser parameters such as fluence, pulse duration, and repetition rate. This molten material can also undergo self-organisation during resolidification, further shaping the surface [29]. For hatching distances below the spot diameter (50 µm) the whole surface area is thermally affected. As a result, formation of micropillars can be observed (figure 3(a)). When the hatch distance exceeds the spot diameter regions between adjacent scan lines remain unaffected by the laser beam and associated thermal effects. As a consequence, isolated parts of unprocessed material persist on the surface forming a distinct square grid pattern (figures 3(b) and (c)). In figure 3, the lines within the laser-processed regions show periodic variations in width rather than uniform tracks. This effect is likely caused by the pulse overlap of 86% in combination with minor synchronisation imperfections between 3 J. Micromech. Microeng. 35 (2025) 105009 K Kobliha et al Figure 1. The contamination application mechanism. (a) For application numerous tools were used (an aluminium sample with laser structures, a squeegee, a measuring cup, and a plate with slits). (b) Impurities were applied to the laser structure using the squeegee and the plate with slits. (c) As a result a defined amount of contamination appeared on the structure. (d) The samples were then placed in an apparatus to measure the self-cleaning effect. the scanning motion and the pulse emission. Consequently, energy is deposited unevenly along the scan path leading to localised material accumulation and variations in line width. To achieve higher contact angles samples were irradiated with a defocused beam with the focal distance from the sample set to 1.7 mm right after laser processing. At this working distance, the beam diameter was approximately 70 µm, corresponding to a laser fluence of 1.30 J cm−2. This additional treatment results in the creation of nanometric protrusions [21] as demonstrated in figure 4. 3.2. Surface wettability characterisation Immediately after processing samples were stored in a vacuum chamber in a high vacuum in pressure 5 ·10−6Pa for 12 h. Subsequently, their wettability was evaluated using 20 µl water droplets. The unprocessed aluminium sample, stored under the same conditions, exhibited a contact angle of (85 ±5)◦and a sliding angle greater than 90◦. Such a surface is clearly unsuitable for self-cleaning, as water droplets adhere to it and are unable to remove contaminants. In contrast, all laser-processed samples exhibited contact angles exceeding 150◦, within experimental error, for all hatch distances in square, rectangular, and rhombic structures (figure 5(a)). Additional irradiation with a defocused beam further increased the contact angle of square and rhombic structures by approximately 10◦(figure 5(b)), whereas the contact angle of rectangular structures remained essentially unchanged. This behaviour can be explained using the Cassie–Baxter model as rectangular patterns exhibited the lowest surface roughness among all tested geometries. Figures 5(c) and (d) shows that almost all measured structures exhibited sliding angles below the superhydrophobic threshold of 10◦, within experimental error. The exceptions are rectangular structures with hatch distances of 35 µm and 45 µm, as well as the 35 µm structure subjected to additional defocused beam irradiation. Additional irradiation with a 1.7 mm defocused beam had minimal effect on the sliding angles of square and rhombic patterns. In contrast, rectangular structures showed a pronounced improvement; for instance, the 45 µm rectangular pattern, which exceeded 15◦ 4 J. Micromech. Microeng. 35 (2025) 105009 K Kobliha et al Figure 2. The experimental apparatus containing 15◦tilted sample in the contact angle measurement device, an external light source, and a camera with the camera lens (a). Desired laser path of different patterns: (b) squares, (c) honeycombs, (d) circles, (e) triangles, (f) rhombuses, (g) rectangles. Figure 3. Topography of square structures for hatch distances: (a) 45 µm, (b) 80 µm, (c) 150 µm and structures with different patterns for hatch distance 100 µm: (d) squares, (e) honeycombs, (f) circles, (g) triangles, (h) rhombuses, (i) rectangles. Sharp peaks are a display error when the microscope sensor does not pick up any signal (typically for dark and deep places) and they do not express the topography in a given place. without treatment, decreased to approximately 5◦following irradiation. This behaviour can be attributed to the initial values of the sliding angles: square and rhombic structures already exhibited low sliding angles (around 5◦) without additional treatment, leaving little room for further improvement. Rectangular structures, however, had higher initial sliding angles, so the effect of irradiation produced a more significant reduction. As the performance of circular, triangular, and honeycomb patterns closely mirrored that of square structures and did 5 J. Micromech. Microeng. 35 (2025) 105009 K Kobliha et al Figure 4. Detail of square structure of hatch distance 80 µm (a) without additional irradiation, (b) with additional irradiation by the defocused beam. Figure 5. Dependence of contact angle (a), (b) and sliding angle (c), (d) on hatch distance for square, rectangular, and rhombic structures without additional irradiation with defocused beam (a), (c) and with irradiation using 1.7 mm defocused beam (b), (d). The lines are for eye guidance only. The portion of the line extending beyond the graph area occurs because the corresponding value exceeds the plotted range (for sliding angles above 25◦), in which case the value is shown as a number with an error. not show any significant improvement in wettability or selfcleaning efficacy further discussion is limited to representative cases. Among all tested pattern geometries the highest contact angle on average can be observed for structures with a hatch distance of 80µm while the lowest for hatch distances of 150µm. As a result, these hatching distances (80µm, 150µm) were selected for subsequent evaluation of selfcleaning performance. 6 J. Micromech. Microeng. 35 (2025) 105009 K Kobliha et al Figure 6. Mechanism of data processing from self-cleaning measurements. 3.3. Self-cleaning properties evaluation Figure 6(a) shows the mechanism for determining the self-cleaning effect. The data was measured on the apparatus described in section materials and methods. The orientation of the laser tracks for the different structures corresponds to figures 2(b)–(g), with the surface tilted in the vertical direction. A region of interest with a size of 350 ×200 px2was selected from the captured photograph of the surface which corresponded to the place affected by a water drop with a diameter of approximately 3.4 mm. The size of the area of interest stayed the same for every surface and particle modification. This area was then converted to black and white using ImageJ with black representing contamination and white representing an uncontaminated surface. In this program the relative area of contamination on the surface was determined as the number of black pixels. The value of contamination at the beginning of the measurement was considered as 1.00 while other values have been correspondingly recalculated to this value. An example of the data processing mechanism is displayed in figure 6(b). Before the passage of the first drop the number of black pixels on the surface was equal to 62 000 px i.e. the area of contamination was approximately 89%. This value was taken as a relative contamination of 1.00. After the application of the first droplet the number of black pixels lowered to value 2350 px i.e. approximately 3%. Therefore, the relative contamination value was calculated as 2350/62000 =0.04. For each additional droplet a similar analysis was applied. In all cases the relative contamination either decreased or remained constant as illustrated in figure 6(b). Using this method originally introduced by Milles et al [1] the evolution of the efficacy of the self-cleaning effect for superhydrophobic surfaces can be determined. The above-described method was applied to evaluate the self-cleaning performance of structures with various shapes fabricated using hatch distances 80 µm and 150 µm both with and without additional irradiation by defocused beam. Figure 7 shows the development of relative contamination as a function of the number of applied droplets. These structures were contaminated using 10 µm sized MnO2particles (figures 7(a), (c) and (e)) or 55 µm sized polyamide particles (figures 7(b), (d) and (f)). As shown in figure 7, regardless of the structural shape or type of impurity, the relative contamination either decreases or remains unchanged as the number of droplets increases. On average the highest cleaning efficacy was observed with a hatching distance 80 µm with additional defocusation. For polyamide the relative contamination after five droplets was approximately 20% while for MnO2it was around 10%. This effect can be attributed to the generally higher contact angles achieved with these parameters compared to other hatch distances and configurations without defocused irradiation (as shown in figure 5). It is known that surfaces with high contact angles and low sliding angles allow droplets to move down with higher velocities [30,31]. Surfaces with high contact angles have less contact between themselves and water droplets and therefore less friction. Surfaces with low sliding angles (and contact angle hysteresis) have lower adhesion and prevent droplets from attaching to surfaces. As a result of these effects higher droplet velocities cause the droplets to move down with nearly constant spherical shapes so the contact area between droplets and surfaces is approximately constant. This enables the droplets to prefer rolling over sliding motion and as a result pick up particles on the surface with high efficacy [1,31,32]. Considering the measured values it can be stated that the self-cleaning properties of the surface are better the higher the contact angle. Regarding self-cleaning efficacy for different types of particles concerning the same type of structural parameters (i.e. hatch distance, shape of structures) more particles of MnO2than polyamide were removed (e.g. figures 7(c) and (d)). Furthermore, relative contamination of MnO2 7 J. Micromech. Microeng. 35 (2025) 105009 K Kobliha et al Figure 7. Relative contamination as a function of number of passed droplets for different types of shapes, hatching distances, use of defocusation, and types of impurities. All displayed structural shapes, i.e. squares (a), (b), rhombuses (c), (d) and rectangles (e), (f) with hatching distances 80µm and 150µm both with and without additional defocused irradiation were contaminated by 10 µm MnO2(a), (c), (e) and 55 µm polyamide (b), (d), (f). The SEM images show the topography of the given shape with the hatching distance 80µm. The lines are for eye guidance only. particles settled down at a value less than 10% after just two droplets whereas relative contamination of polyamide particles decreased after each droplet (e.g. figures 7(a) and (b)). This could be explained by the size (and consequently weight) of the particles when smaller and lighter 10 µm MnO2 particles were removed with ease as water droplets bound these particles on their surface better than larger and heavier 55 µm polyamide particles. Additionally, the intrinsic affinity of the particles can play a role: MnO2is hydrophilic [33], which enhances its adhesion to water droplets, whereas polyamide tends to be hydrophobic [34], reducing droplet-particle binding. Regarding the influence of structural shape on the selfcleaning effect rhombic patterns demonstrated the best performance for both types of impurities (figures 7(c) and (d)). In most combinations of hatch distances and defocused irradiation the relative contamination of MnO2dropped to 0% (within the measurement error) just after three droplets (figure 7(c)). Similarly, polyamide contamination was the lowest among all tested shapes for rhombic structures with residual contamination levels around 15% after five droplets in most cases. An exception was observed for the 150 µm hatch distance without defocused treatment where the remaining contamination remained at 40% (figure 7(d)). 8