A comparison of characteristics of periodic surface micro/nano structures generated via single laser beam direct writing and particle lens array parallel beam processing

ABSTRACT


INTRODUCTION 70 71
Modifying surface properties by producing tailored nano/microstructures have 72 been extensively died and has found wide applications in several fields including self-73 cleaning, coating adhesion, wear resistance, anti-icing, and biological applications. 74 Moreover, changing the surface optical properties is used for solar cell and photoresist 75 applications [1][2][3][4][5][6][7][8]. Many surface engineering methods are available to modify the 76 surface structures, including chemical etching, sandblasting, mechanical machining, 77 corona discharge, plasma etching, and laser surface texturing. Among these techniques, 78 laser surface texturing is one of the most efficient and flexible techniques. It has been 79 used to produce smart surfaces that meet the requirements for more stringent surface 80 structure designs and properties. Besides, laser surface texturing has the advantages of 81 fast processing speed, non-contact processing, ease for automation, zero tool wear, and 82 non-dross ablation [9,10]. 83 Different techniques have been used in laser surface texturing for producing 84 different structures, including laser direct writing with or without using a mask, two or 85 multi-beam interference and particle lens array multiple laser beam patterning [11]. 86 According to the surface feature size, laser surface texturing can be classified as micro-87 scale, nano-scale, or multi-scale surfaces (hierarchal structures, with a combination of 88 micro/nano features). Various structures such as holes, grooves, bumps, protrusions, 89 periodic surface structures, spikes, conic structures have been produced using laser 90 direct writing [9,12,13]. The generation of various types of surface patterns depends 91 mainly on the laser parameters, such as including fluence/power density, wavelength, 92 This work was motivated by the practical engineering challenges of producing surfaces 115 with various micro and nanostructures using flexible and efficient ways. Therefore, in 116 this paper, two laser surface texturing techniques are compared to examine the 117 properties of surfaces with various structures and properties. The first technique is the 118 laser direct writing. This process is performed using nanosecond laser processing of a 119 stainless steel surface to modify the surface structure and properties. The second 120 technique is the contact particle lens array used to generate micro/ nanostructures on a 121 thin film of GeSbTe (GST) film coated on a polycarbonate substrate. 122

EXPERIMENTAL PROCEDURE 123 124
For the laser direct writing method, AISI 316L stainless steel sheets were used in 125 this work. These sheets were cut to the dimensions of 10 mm x 10 mm x 0.7 mm. Before 126 the laser treatment, the sheets were cleaned ultrasonically using acetone, ethanol, and 127 di-ionized water, for 10 minutes each and dried used compressed air. 128 A Nd:YVO4, nanosecond laser (Laserline Laserval Violino) was selected to study 129 the effect of laser processing parameters such as scanning speed and hatch distances on 130 micro structuring wettability, reflectivity and oxygen surface content. The effect of 131 scanning direction, laser fluence, laser scanning speed and scanning environment on the 132 surface morphology were examined. The laser beam was directed using a set of x-y 133 Galvo scanning mirrors and an F-theta lens. In some experiments, the sample was 134 submerged in water in order to reduce surface oxidation and reduce the feature sizes 135 and compared that processed in air. The level of water for the submerged sample 136 experiments was 1 mm above the sample. The laser scanning was performed in one direction, two or more directions. The effect of scanning direction on the surface 138 morphology was studied. The experimental scheme is shown in Fig. 1. Table 1 listed the 139 laser processing parameters used in this work. 140 Prior to the surface characterization after laser treatment, the samples were 141 ultrasonically cleaned using Ethanol then compressed air to remove any ablated 142 materials and contamination. Scanning electron microscopy (Philips XL30 FEG-SEM) was 143 used to examine the surface morphology. This SEM is combined with energy dispersive 144 X-ray (EDX) which was used to characterize the surface oxygen contents. A confocal 145 laser scanning microscope (type: Keyence 3D profiler) was used to examine the surface 146 roughness. Water drops sessile method using a contact angle analyzer (type: FTA 188) 147 was used to investigate laser textured surfaces' wettability characteristics. In this 148 method, a contact angle of 10 µl droplets of de-ionized water that contact the surface 149 was measured. 150 For particle lens array experiment, GeSbTe (GST) film coated on a polycarbonate 151 substrate was used in this work. GeSbTe is a composite of three materials (germanium-152 antimony-tellurium or GST) and it is a phase-change material from the group 153 of chalcogenide glasses used in rewritable optical discs and phase-change 154 memory applications. The crystallization temperature of this alloy is between 100 and 155 150 °C, and its melting point is around 600 °C. It is characterized as a high speed phase 156 change material and its crystallization time around 20 nanoseconds which make it easy 157 for patterning. The reason for the selection of this material is because it requires very 158 low laser energy density to cause surface morphology change, ideal for the particle lens surface patterning applications. The sample was a 20 nm thick Germanium-Antimony-160 Tellurium (GST) film coated on a polycarbonate substrate. A low concentration of 4.74 161 µm SiO2 microspheres was prepared by buffering the microspheres solution in de-162 ionized water. The solution was spread over the substrate to form a monolayer. Then 163 the water evaporated when the samples were placed with a 90° angle. 164 A KrF Excimer laser (GSI-Lumonic IPEX848) was used to irradiate the samples using 165 the setup shown in Fig. 2. The laser beam size was 25×25 mm with a uniform intensity 166 distribution and a lens (focal length = 10 mm) was used to focus the laser on the sample 167 with a spot area 10 mm × 10 mm. the laser processing parameters are listed in Table 2. 168 For imaging the microsphere morphology formed on a film substrate and nano patterns 169 generated after laser irradiation, optical microscopy (type: Leica CH-9435) was used. Various and homogeneous structures ranging from ripples to porous and conical 175 structures were achieved by changing the scanning directions, scanning speed, laser 176 fluence, and processing environment. Figure 3 shows the nanosecond laser surface 177 texturing's typical surface morphology using five different scanning patterns. It can be 178 seen that by changing the angle of scanning from (0˚, 90˚) to (10˚, 130˚), (0˚,90˚,45˚, -179 45˚), and (10˚, 130˚), the surface morphology was changed from conical like structure to 180 micro-pores, beehive, and diamond-like structures, respectively. The scanning was performed at a 100 mm/s scanning speed, a 50 µm hatch distance, a 7 ns pulse 182 duration, a 30 kHz pulse repetition rate, a 3.9 J/cm² laser fluence and 10 repeat scanning 183

passes. 184
The effect of laser scanning speed on the microstructure morphology was also 185 studied (Fig. 4). The scanning was performed using one direction scanning and laser 186 parameters of (9.8 J/cm² laser fluence, one pass, and 50 µm hatch distance). It can be 187 seen that changing the scanning speed from 2000 mm/s to 1 mm/s changed the 188 microstructure from an oval-like structure to a cauliflower-like structure. 189 Figure 5 shows the microstructure of surfaces generated using three laser 190 fluences. The scanning was performed using two directions (30˚ and 60˚), using laser 191 parameters of (100 mm/s scanning speed, 50 µm hatch distance and 10 times passes). It 192 can be seen that micro ripples were formed using low laser fluence (0.9 J/cm²) (Fig. 5a),193 while it changed to a hierarchal microstructure using a high laser fluence (9.8 J/cm²) 194 (Fig. 5c). 195 Figure 6 shows the difference between the microstructures generated in water 196 and air. Using laser parameters of (9.8 J/cm², 10 mm/s scanning speed, 25 µm hatch 197 distance, and 1 pass), the microstructure in the air was cauliflower likes structure, while, 198 in water, a uniform conical like structure was formed. 199 Changing the scanning direction and scanning speeds affects the number of 200 pulses and the laser overlapping which in turn affected the surface structure. The pulse 201 overlapping can be calculated as [9,21]: 202 The line overlapping can be calculated as [9,21]: 204 (2) 205 Where, ∆ = and ∆ is the hatch distance. 206 In this work, the line overlapping was 81.8 %, and the pulse overlapping was 39 207 % and 99.9 % at 1000 mm/s and 1 mm/s laser speeds, respectively. The number of 208 pulses per spot can be estimated as * , and it was 2 and 1650 at a 209 laser speed of 1000 mm/s and 1 mm/s, respectively. 210 By increasing the laser fluence from 0.9 J/cm² to 9.8 J/cm², the surface 211 microstructure changed from micro ripples to a 3D complex structure. This is related to 212 the increase the material removal by increasing the laser fluence. The removed 213 materials could be solidified around the laser scanning area forming complicated 3D 214 structures [22]. 215 By changing the laser-processing environment from air to water, the 216 microstructure was changed to be smooth and free of solidified material and particles 217 over the microstructure. The reason behind this is that the ablated particles were 218 moved with water movement and they did not redeposit over the surface during laser 219 processing of the material in water [9,23,24]. 220

Surface characteristics 221
The surface roughness (Ra) measurements of as-received surface (control) were 222 10.9±3.54 nm. Figure 7 shows the roughness values of laser treated surfaces using a 223 range of scanning speeds and hatch distances. It is clear that the surface roughness was 224 proportional to the hatch distances and inversely proportional to the laser scanning 225 speed. The surface generated using a hatch distance of 100 µm and a scanning speed of 226 10 mm/s recorded the highest surface roughness Ra value (11.12 ± 1.8 µm) compared to 227 those of other surfaces. Surface generated using a 10 µm hatch distance and a 1000 228 mm/s scanning speed, on the other hand, showed the smallest surface roughness (0.12 229 ± 0.007 µm) compared to other surfaces. 230 Figure 8 shows the analysis data of the energy-dispersive X-ray spectroscopy 231 (EDXs) for surfaces. As received surface (control) was free of oxidation as its' measured 232 oxygen percentage was zero. However, after laser processing, it was noticed that the 233 surface oxygen content of all processed samples was increased. By increasing the laser 234 scanning speed and the hatch distance, the oxygen percentage decreased. For example, 235 at a speed of 10 mm/s, the oxygen percentage recorded 21.8 % and 7.8 % using, 236 respectively, 10 µm and 100 µm hatch distances. However, at a speed of 1000 mm/s, 237 the oxygen percentage was 0.93 % at 10 µm hatch distance and less than 0.4 % using 238 100 µm hatch distance. 239 In this work, the effect of ns laser generated surface structures on the change of 240 the stainless steel (SS) wettability was investigated. The contact angle of as received 241 substrate (control) was 90.5˚±3.5˚. After laser treatment, all the surfaces performed 242 superhydrophilic properties with a contact angle CA=0˚ immediately after the laser 243 processing. However, the wettability characteristics of all processed samples changed 244 with time. Therefore, the contact angle was measured again at one month after laser 245 processing. Figure 9 shows the surface wettability change as a function of the laser 246 scanning speed and hatch distances. It is clear that the contact angle increased with 247 increasing scanning speed and hatch distance. The surface produced at a 1000 mm/s 248 and 100 µm hatch distance was superhydrophobic with maximum contact angles (CAs) 249 around 158°. However, the minimum CA was ˷ 125° for the surface generated at a 10 250 mm/s scanning speed and 10 µm hatch distance. Generally, the wettability of surfaces 251 was inversely proportional to the scanning speed and hatch distance. 252 Figures 10 and 11 show the change of surface reflectivity within the visible light 253 spectrum (400 -700 nm) after the laser processing. The reflectivity was investigated for 254 samples treated using different laser speeds and hatch distances. It can be seen that the 255 reflectivity of all processed surfaces was decreased compared to the reflectivity of as 256 received substrate (control) (Fig. 10). Furthermore, at a specific hatch distance, it can be 257 seen that the reflectivity of surfaces was increased with increasing the scanning speed 258 (Fig. 10). Moreover, at a specific scanning speed, the results show that the reflectivity of 259 laser-treated surfaces was increased with increasing hatch distance (Fig. 11). The 260 reflectivity of as received stainless steel (control) was about 60 %. However, the 261 reflectivity was decreased to less than 2 % for samples treated using 10 mm/s laser 262 speed and 10 µm hatch distance. 263 In this work, it was observed that laser processing parameters affected the 264 surface properties. Changing the surface characteristics with changing the laser 265 processing parameters has been extensively studied before. Some researchers have 266 reported that the surface wettability increases with increasing the surface oxygen 267 contents and surface roughness [25] . With increasing surface roughness, the contact area between the surface and water droplets increased due to the natural gravitational 269 desire to settle on the surface [26,27]. Cui et al. [28] reported that heating the stainless 270 steel increased the oxygen contents from the surrounding environment forming Fe2O3 271 and Cr2O3. The hydroxyl group (OH density) increased the surface oxygen contents, 272 thereby increasing the surface adsorption and surface wettability [26,27]. In our work, 273 during material processing, it was noticed that the surface reflectivity was decreased, 274 and some surfaces switched to black. Other researchers also noticed this behavior 275 during different processing materials such as Si [29]. This is related to increasing the 276 surface roughness, which leads to increased surface area and multiple reflections inside 277 the surface features [9,30]. 278

Particle lens array parallel laser beam surface patterning 279
Parallel processing techniques are essential to generate micro/nano-textures 280 over a large area. Particle lens arrays can be used to produce micro and nanostructures 281 by splitting a single laser beam into millions or billions of laser beams and focusing 282 locally without diffraction limit. This technique is based on near-field effect of small 283 transparent microspheres to produce micro/nano-patterns. Using this technique, it is 284 possible to produce a feature size below the diffraction limit. Monodispersed 4.74 µm 285 spherical silica (SiO2, Duke Scientific) particles were diluted with de-ionized water and 286 applied to the film surface. After the water evaporated, a hexagonally closed-packed 287 monolayer was formed on the surface due to the self-assembly process. The sample was 288 then laser processed and characterized using a Leica CH-9435 microscope (Fig. 12). Motion Controllers/Drivers (due to their common commands and answers) and can be 304 easily adjusted for other motion control systems and lasers according to the forward 305

documentation. 306
User-defined complex shapes can be fabricated within regions dp ≤ r, as shown in 307 angle φ moves p in circumferential direction. The ranges of angles are α(-45°, 0°) and φ(-180° , 180°). By applying a relative angle α(φ) with every rotated angle φ, user-defined 311 patterns can be easily fabricated. 312 According to the Mie theory, the induced near-field enhancement is located 313 around the particle and along to incident direction. It is known that the enhanced 314 intensity will decay along the incident direction before it reaches the substrate surface. 315 If the laser energy is sufficient, this enhanced field is still able to ablate materials for the 316 substrate surface. The shift of these peak positions away from the contact point with a 317 distance close to that is given by the geometrical optics: 318 where r is the radius of the particle. 320 Figures 14 and 15 show two periodic patterns generated on the GeSbTe film 321 using an excimer laser. Figures 14a and 15a show the computer design of a square 322 shape and (nano) in the software interface, which led to the patterning of the GeSbTe 323 (GST) film with uniform periodic patterns. The gaps between dots cannot be 324 distinguished due to their overlapping (one can see in Figs. 14a and 15a) and of melting 325 of the film during the laser processing. The fluence used was 1 mJ/cm 2 . The experiments 326 were performed with two Newport PR 50 Series computer-controlled rotation stages 327 and a Newport ESP300 controller. One stage controls the laser incident angle α by tilting 328 the sample with angles ranged from 0° to 45°. The other stage rotates the sample within 329 the tilted plane with an angle φ with angles ranged from -180° to 180°. Therefore, any 330 point p (α, φ) within the shade of particle on the substrate surface could be reached by 331 a geometrical calculation of angles α and φ.
The developed technique could provide means to produce arbitrary patterned 333 surfaces on small objects such as MEMS (for improved tribological characteristics), 334 OLEDs (for improved emission efficiency), optical metamaterials, uniform structures for 335 cell and bacterial adhesion and migration, and medium-sized objects. 336 In this work, the results showed that the surface properties and structure could be 337 controlled by controlling the processing method. Using laser direct writing method, the 338 results showed that increasing the surface wettability and absorptivity were related to 339 increasing the surface roughness and oxygen percentage content. However, the 340 controllability of the type of surface micro/nano patterns is limited. The parallel laser 341 beam processing using a particle lens array, on the other hand, allowed rapid production 342 of user-designed periodic surface patterns at nano-scale, overcoming the optical 343 diffraction limit with a high degree of controllability where controlling the uniformity of 344 the particle lens array is a challenge. 345

CONCLUSION 346
In this work, nanosecond pulsed lasers were used to generate different 347 micro/nanostructures using two different techniques: direct writing and particle lens 348 array parallel writing. In the laser direct writing technique, the substrate was melted and 349 evaporated, and then the re-deposition and solidification of molten materials generated 350 different microstructures. Therefore, the formation of different structures using the ns 351 laser was mainly due to the thermal effects on treated surfaces. The surface 352 morphology and properties were changed with changing laser-processing parameters. 353 Moreover, the surfaces processed a low scanning speed (10 mm/s) recorded the highest 354 roughness and oxygen percentage content and the minimal wettability and reflectivity compared to other surfaces. Furthermore, the surfaces produced at 100 µm recorded 356 the highest roughness, water contact angle and reflectivity and minimal oxygen 357 percentage contents compared to other surfaces generated using 10 µm and 50 µm 358 hatch distances. Thus, the surface of high roughness and oxygen percentage content 359 presented a high wettability and turned to black with low reflectivity characteristics. 360 Using particle lens array technique, two nano-patterns were demonstrated. This is a 361 very efficient way of producing tailored surface micro/nano patterns. Both processing 362 methods presented effective surfaces. This study shows that nanosecond lasers could 363 generate distinctive morphologies and properties using easy and efficient ways. These 364 surfaces could be useful in various applications, including biological applications, wear 365 resistance, self-cleaning, anti-icing, coating adhesion, and solar cells .  366  445  446  447  448  449  450  451  452  453  454  455  456  457  458  459  460  461  462  463  464  465  466  467  468  469  470  471  472  473  474  475  476  477  478  479  480  481  482  483  484  485  486  487  488  489  490  491  492  493  494   The surface roughness measurements of laser treated surfaces Fig. 8 The surface oxygen contents measurements of laser treated surfaces Fig. 9 The contact angle measurements of laser treated surfaces   Table Caption List  770  771  Table 1 The used nanosecond laser parameters for texturing the stainless steel Table 2 Laser parameters for inducing different surface textures using Excimer laser