Direct Laser Processing of Two-Scale Periodic Structures for Superhydrophobic Surfaces Using a Nanosecond Pulsed Laser

A steel (PSL, Ra = 0.26 (n = 24) of surface roughness, 33–37 HRC of hardness, Hitachi Metals Tool Steel Ltd., Tokyo, Japan) was used as the material fabricating a square-shaped test-piece (test-piece for wettability evaluation, 20 mm length × 20 mm width × 5 mm thickness) to produce two-scale periodic structures on the surfaces. Because of the difficulty of cutting a thin rod, a pin-shaped steel (SKD-11, 58.0 HRC of hardness, Hitachi Metals Tool Steel, Ltd.) was instead used for a pin-shaped test-piece (test-piece for depth evaluation, 1.5 mm diameter) to estimate the depth of the percussion drilling. Table 1 shows element compositions of steels used for fabricating test-pieces.

The processing of a two-scale periodic structure using a nanosecond pulsed laser is shown schematically in Fig. 1. The two-scale periodic structure consist from (i) a basically periodic surface structure formed by the pitch of the percussion drilling and (ii) a small-scale periodic surface structure formed by the molten material around the micro-holes. A commercial nanosecond pulsed laser system (ML-9011A, 10 W, YVO₄ solid-state laser, Amada Miyachi Co., Ltd., Isehara, Japan) that provides an oscillation wavelength of 532 nm, a pulse width of 11 ns, and a repetition rate of 20 kHz was used for the experiment (Fig. 1a). The laser beam with a Gaussian profile was focused perpendicularly on the test-piece using a beam expander (× 8, Geomatec Co. Ltd., Kanagawa, Japan), using an fθ lens (focal length: f = 160 mm, Geomatec Co. Ltd., Kanagawa, Japan) and two-dimensional galvanometric mirrors. The spot diameter at the focal point on the test-piece surface was calculated as 16 μm by 1/e² using a laser beam quality factor of M² = 1.2 and an expanded laser beam diameter of 8 mm [31].

The laser scanning route consisted of parallel beam lines, the distance (pitch) between which was used as a parameter, τ (τ_x = τ_y, Fig. 1b). The laser scanning velocity, v_s, was set to 1500 mm/s for both directions, achieved using the two-dimensional galvanometric mirrors. The nanosecond pulsed laser was used for percussion drilling so that both thermal and ablation processes would occur (Fig. 1c). Consequently, a two-scale periodic structure is formed by processing two periodic structures simultaneously (Fig. 1d).

Twenty four test-pieces for wettability evaluation were prepared using the pitch (16–120 μm), number of repetition shots, s (1–120), and fluence, F (2.49–20 J/cm²), as the parameters (Table 2).

The equilibrium contact angle, θ, apparent contact angle, θ’, and contact angle hysteresis (CAH = θ_rec – θ_adv) of the test-pieces for wettability evaluation were measured using a commercial contact angle analyzer (DM-701, Kyowa Interface Science Co. Ltd., Niiza, Japan) by dropping a droplet of distilled water from a microsyringe. The volume of droplet was 2 μL for measuring θ and θ’. The volume of droplet was 30 μL for measuring CAH. The measurements of θ and θ’ were repeated for five times (n = 5), and the mean values were used. For CAH, time-course changes of CAHs were measured, and the mean values of five data points (n = 5) were used.

It has been reported that wetting behavior is related to the amount of organic compounds on the surface with time [32‐34]. Therefore, the time-course changes of the apparent contact angles were measured for a month and a half period to capture any changes. Immediately after the laser processing, the test-pieces were kept at 48 °C in an incubator (MIR-153, PHC Holdings Co., Tokyo, Japan) during this evaluation to accelerate the oxidation aging.

Figure 2 shows a time chart of the nanosecond pulsed laser used for this laser processing. The processing time per shot, t, is shown by the sum of (i) the percussion drilling time, t_p, (ii) the time taken to move the scanner to the next position, t_m, and (iii) the scanner settling time, t_s. as follows:

The percussion drilling time, t_p, is given as follows:where, s: Number of irradiations (= 1–120),

In this experiment, t_p was between 0.05 and 6 ms. The maximum value of t_m (= τ/v_s) was estimated as 0.08 ms from the laser scanning velocity (v_s = 1500 mm/s) and the maximum pitch (τ = 120 μm). For the present nanosecond pulsed laser system, t_s was 1 ms, in which case t_m is negligible. Thus, the total processing time, T, is expressed in terms of the number of holes, N, as follows:When both the pitch (τ_x = τ_y) and the length of the processed area (l_x = l_y = l) are the same in both directions, the number of holes per unit area, N, is given as follows:

Thus, the processing rate, v, is given as follows:

Figure 3 shows a sequence of laser confocal microscope micrographs of the test-pieces for wettability evaluation for various laser fluences (samples no. 21–24 in Table 2). As expected, crater-shaped holes were observed in the direction of the laser beam. It was considered that the bank-shaped outer rim around each micro-hole was formed by the molten material which was squeezed out of the micro-hole by the expanding plasma. Mirza et al. reported that craters could be processed by using a femtosecond pulsed laser, however the heights of the outer rim did not change significantly [23]. In Fig. 4, the structure processed on steel by percussion drilling is illustrated schematically as a two-scale periodic structure, where f₁ is the width of the solid–liquid interface, f₂ is the width of the liquid–air interface, h₁ is the depth of the basically periodic surface structure, f_s1 is the width of the small-scale periodic surface structure, and h₂ is the bank height of the small-scale periodic surface structure. The values of f_s1 could not be determined under the condition of 16 μm pitch due to the overlap of two-scale periodic structures. The value of f_s1 was in the range of 2.9–10.2 μm. The values of the bank height, h₂, for fluence values of 2.94, 4.98, 7.46, and 9.95 J/cm² were 0.7, 2.8, 6.0, and 8.2 μm, respectively. Therefore, we reason that the bank-shaped outer rim was caused by the molten material generated by the expanding plasma of the nanosecond pulsed laser, meaning that the bank height could be controlled when using a nanosecond pulsed laser.

The results for the geometrical parameters of the two-scale periodic structures measured using the non-contact laser confocal microscope are shown in Table 3. The value of h₁ at τ = 16 μm could not be measured because adjacent processing areas interfered. For sufficient pitch length, h₁ was considered to be almost constant and independent of the pitch. The f₁/τ ratio was in the range of 0.37–0.87, increasing in proportion to the pitch. Thus, the f₂/f₁ ratio was in the range of 0.15–1.69, decreasing in proportion to the pitch. The diameter, f₂, changed with the pitch distance and the number of laser pulses. It was considered that the molten material was squeezed around the micro-holes in proportion to the number of laser pulses, whereupon f₂ decreased.

Figure 5 shows the results for the hole depth, h₁, measured using the test-pieces for depth evaluation and three-dimensional X-ray microscopy because the aspect ratio of the holes prevented them from being measured fully using the laser confocal microscope. Figure 5a and b show how h₁ varied with the fluence, F, and the number of repetition shots, s, respectively. The value of h₁ increased linearly in proportion to the fluence (Fig. 5a). The absolute values of h₁ for the SKD-11 steel were similar to those of the PSL steel for s = 1 and 20 (Table 3). The compositional differences between the two steels were those regarding C, Ni, V, and Cu, clarified as being a few percent. Thus, it was considered that the ablation thresholds of the two used steels comparable to lead to the same surface features. A limitation was that the single pulse results should be excluded because of the resolution of the micro-CT. The hole depth could potentially increase exponentially because reflectivity decreases with increasing temperature [35]. However, the value of h₁ also increased linearly in proportion to the number of repetition shots (Fig. 5b). Therefore, we assumed our conditions limited the increase in surface temperature sufficiently so that we can use both the fluence and the number of repetition shots to estimate the hole depth in a linear manner. These data were used to estimate the processing rate for the two-scale periodic structures.

Figure 6 shows the time-course changes of the apparent contact angles immediately after the laser processing of the test-pieces for wettability evaluation (samples no. 1–20 in Table 2). The apparent contact angles of the steel surfaces increased over time, and the surfaces became nearly superhydrophobic. Jagdheesh et al. reported that laser processing of metal surfaces creates preferential sites for the adsorption of organic compounds from the air [26]. The change in wetting behavior was correlated with the amount of carbon on the structured surfaces [33]. It was considered that the apparent contact angles to have stabilized after 4 weeks and we used those values to evaluate the hydrophobicity.

The maximum contact angles after aging for each test-piece for wettability evaluation are shown in Table 3. The measured apparent contact angle increased as the pitch increases from 16 to 80 μm, after which it decreased. The maximum apparent contact angle of 161.4° was associated with a pitch of 80 μm and 20 repetition shots. This condition might be the optimum condition based on the Cassie-Baxter equation. This superhydrophobicity was equal to or above with the previous report of the hierarchical structure [36]. Also, the same conditions are associated with the minimum contact angle hysteresis of 4.2°. These results agreed well with those in a previous report that found the hysteresis values to be less than 10° when the apparent contact angle exceeded 150° [37].

It seems that the contact angle did not change much after a few pulses (Table 3). In particular, the contact angle for s ≥ 20 remained almost constant for pitch values of 40 and 80 μm. These findings suggested that we could focus on using low pulse numbers. Reducing pulse numbers contribute to reduce processing time and heat affected zone. If the number of repetition shots increases, unstable morphology may appear in the pores, which may not be conducive to obtaining a stable superhydrophobic structure.

Table 4 shows the relationship between geometrical parameters and wettability of the two-scale periodic structures. In this table, the depth of the basically periodic surface structure, h₁, was estimated by using the linear regression analysis between the hole depth and the fluence (Fig. 5a). The h₁ saturated when fluence exceeded 80 J/cm². It was considered that the increase in the hole depth slowed down by the plasma shielding effect [38]. The apparent contact angle showed hydrophobicity when the basically periodic surface structure, h₁, and the bank height, h₂, exceeded a threshold level. The absolute values of each threshold level will be changed by geometrical parameters such as pitch.

The values of the apparent contact angle, θ’, were calculated based on the Cassie-Baxter equation and introducing a non-dimensional roughness factor R_f (≥ 1) on surface f₁ as follows [39, 40]:where θ is the equilibrium contact angle, and is equal to 97.3 ± 3.2° (mean ± standard deviation, the mean value was calculated using 24 test-pieces for wettability evaluation). This theoretical model adapts the hypothesis that the water is under zero hydrostatic pressure. This model approximates the two structures as a basically periodic surface structure and a small scale structure on surface f₁. Figure 7 shows the relationship between the apparent contact angle and f₁/τ ratio for these structures. The measured apparent contact angle decreased with increasing f₁/τ ratio. The calculated results agreed well with the measured results when R_f = 1–7 were used. These results confirmed that the present theoretical model pertains to the two-scale periodic structures.

Figure 8 shows the processing rate of the two-scale periodic structures calculated using Eq. (5). Under the present conditions, the processing rate ranged between 2 and 823 mm²/min, increasing with the pitch and decreasing with the number of repetition shots. In this method, the number of repetition shots can be reduced by increasing the fluence. Eq. (5) showed that increasing the repetition rate, f, was an effective way to increase the processing rate. Consequently, a maximum processing rate of 823 mm²/min was achieved with τ = 120 μm and s = 1. This processing rate equivalent to 28.7 mm square per min and/or 222 mm square per hour. In previous research involving percussion drilling using a nanosecond pulsed laser, Cai et al. used an area of 10 mm square area, 110.9 μm pitch, 1 shot/hole, and 0.4 s duration time on stainless steel [41]. Yang et al. used a condition of 20 mm diameter area, 100 μm pitch, 24 shots/hole, and 50 mm/s scanning speed on nickel-chromium-based superalloy [42]. Although neither report states the processing rate clearly, we estimated that it was in the range of 1–100 mm²/min. Therefore, the present results supported the assertion that a rapid fabrication method has been obtained, leading to fast laboratory-to-fabrication transfer. With superhydrophobicity, we can use s ≥ 20, in which case the processing rate will be 432 mm²/min with τ = 120 μm. A stop-and-go action is necessary to increase the roundness of the work marks on the workpiece. However, if such stop-and-go action can be avoided and the pulsed laser source moved continuously, then the throughput can be increased substantially.