Influence of Pulse Energy and a Material’s Magnetic Field on the Morphology of the Dimples Produced with Nanosecond Pulsed Laser on a Magnet

Magnets made of NdFeB alloys are applied in numerous industries because they have excellent magnetic properties and good thermal stability. However, their low corrosion resistance and fragility reduce the number of the applications in harsh environments. Laser surface texturing is a powerful technique to improve the properties of materials via surface functionalisation. The characteristics of the textures define the properties of the functionalised surfaces and the laser parameters in turn influence the texture features. Pulse energy is a laser parameter with high influence on the geometry and the morphology of the textures. The magnetic field of these magnets are also expected to have an influence on the texture features. In this present work, the influence of the pulse energy (from 5µJ to 696µJ) on the dimple characteristics created with an infrared (1064 nm) nanosecond pulsed (200ns) fibre laser single pulses on NdFeB alloy (magnetic material) have been studied. The shape features of the laser textures on the samples were analysed through optical microscopy and profilometry. This work shows that the single pulses produced textures of U-type dimple kind on the metallic material. The shapes of the dimples were also altered by magnetic field of the material, which indicates that the magnetic properties of the alloys remain in liquid state. The factors of the laser beam-material interaction, pulse energy and energy depth penetration, are furthermore presented in this paper. These factors can be employed to design the width, diameter and depth of the dimples.

Surface texturing is a treatment that can improve a number of properties via the functionalisation of the surfaces. This process can be carried out by mean of various methods; e.g., electromechanical machining, chemical etching, plasma, laser and ion beam. Laser Surface Texturing (LST) stands out among them because of its advantages such as, fast, environmentally friendly, automatable, high reproducibility and accuracy [8][9][10].
The properties of the processed surfaces are defined by the texture characteristics (e.g. width, diameter, depth and average roughness) [11,12]. These features are, in turn, determined by the laser parameters (wavelength, pulse length, pulse numbers per shot and pulse energy) and the optical and thermodynamic properties of the material [11,13]. The textures on metallic materials are produced by thermal process for lasers with infrared (IR) wavelengths (λ). The laser radiation is absorbed by free electrons that subsequently collide with lattice phonons. These collisions convert the absorbed laser energy to heat, which can melt and vaporise the material [14,15]. Then, the laser treated zone is rapidly cooled by the thermal diffusion of the heat in the material. This is the process responsible to generate the textures [10]. In addition, the heat diffusion into material causes a damaged area round the texture that is often called as Heat Affected Zone (HAZ) and commonly impoverishes the texture quality [16,17]. HAZ thickness is proportional to the square root of the pulse length of the laser (τ) multiplied by the thermal diffusion of the material (κ). The nanosecond pulsed lasers are commonly employed in LST because they produce small HAZ [17]. The amount of material vaporised by the laser beam is highly influenced by pulse energy (E p ). The size of the texture is commonly proportional to vaporised material amount and therefore, it also is proportional to E p [18]. However, a plasma can be generated at high Ep and short pulse length, τ, which can reduce the LST efficiency. Part of the laser radiation can be absorbed by evaporated material that can become plasma at certain values of E p and τ. The plasma can absorb, scatter and reflect the laser beam that arrives on the surface of the material. This diminishes the laser radiation amount that reaches on surface and therefore the energy absorbed at the surface is reduced. This effect is usually named as plasma shielding effect [19][20][21]. Single shot mode is recommendable for texturing because the probability of the plasma shielding effect generation is increased with an increase of the number of pulses hitting a point on the surface (N) [21,22]. Thus, the pulse energy, E p , of the single shots is a great important factor on LST.
Note, the NdFeB alloy is a magnetic material whose magnetic field can disturb the quality of the texture. The dispersion of the vaporised material and the hydrodynamic movements of the molten material define the features of the textures. The dispersion and the liquid movements can be hindered by the magnetic field of the solid material because it produces opposite forces to direction to these processes. This is due to the interaction of the magnet's field and the vaporised and molten metallic alloys, [23][24][25].
Although several researchers have studied the influence of the laser parameters on texture features for metallic materials [26][27][28], the influence of E p on texture characteristics on NdFeB alloy has scarcely been reported in the literature. For this reason, the present work is a study about the influence of E p on the texture features generated via single pulses of a nanosecond pulsed (200ns) IR (1064 nm) fibre laser on a NdFeB alloy magnet. The texture quality is also analysed in this study to assess the influence of the magnetic field of the NdFeB alloy on the laser surface texturing process. The textures produced with single shot were dimples that were analysed through optical light microscopy and profilometry. The study showed that the geometry of the dimples on NdFeB alloy had acceptable quality. The dimple shape was U-type [29][30][31][32]. Moreover, the factors of the laser beam-material interaction (energy depth penetration, pulse energy and energy fluence threshold) have been measured in this study. These factors allow selection of laser parameters to create a dimple of a given diameter and depth.

Materials and pre-treatment
The magnet used is a N42 neodymium-iron-boron (Nd-Fe-B) magnet, type F646-1 supplied by Magnet Expert. The size of the samples was 20 mm diameter and 10 mm thickness. The chemical composition of the Nd-Fe-B is summarised in the Table 1.
The surfaces of the samples were polished before the laser processing (pre-treatment). The polishing process was carried out in three successive steps. First step was a subsequent grinding process with silicon carbide papers of P400, P600 and P1200. The next step was a sequential polishing via polycrystalline diamond pastes of 6 μm, 3 μm and 1 μm. The last step was a mirror polishing using a colloidal silica gel dissolution that consisted of 50% in volume of silica gel (0.04 μm of grain size) and 50% in volume of distilled water. All polishing consumables and items supplied by Struers. Note a cleaning process was conducted after each polishing process. The cleaning treatment was comprised of a first cleaning with commercial detergent and rinsing with fresh water, then an isopropanol spraying and, finally a fast drying with dryer.
The laser beam was addressed to the sample surfaces with the following optical system: The beam expander widened the diameter of the laser raw beam (D R ) to 5.7 mm and also guided the laser beam to scanning galvanometer mirror. The laser beam was addressed to the f-theta focussing lens via the scanning galvanometer mirror. This device controlled the displacement of the laser beam on the sample surfaces and the scan rate of the laser beam (SR) could be varied from 0.01 mm/s to 20,000 mm/s. The focal lens had a focal length (FL) of 100 mm. Note, the samples were placed on the three axes automatic table and the beam expander was manually set.
In this study, the laser device was working in pulsed mode with τ = 200ns τ, PF = 25 kHz and with straight line scanning. The laser beam was focused on the surfaces of the samples. The theoretical laser beam diameter (d o(Theo) ) was 51 μm, calculated using Eq. 3 [26,27,42] and the values of the Table 3 for each parameters. This value was estimation form theoretical values.
The laser was scanned over the sample surfaces to achieve single pulse exposures.
The distance between pulses (d t ) was set to avoid any possible overlapping of the pulses, which was 300 μm. The SR was 7500 mm/s, calculated using Eq. 4: E p was varied from 5µJ to 696µJ in this work. E p were estimated with the Eq. (5) [41] and the average power (P ave ) measurements after galvanometer mirror scanner.
Maximum E p was lower than E pm because the scanning galvanometer mirror reduced the power at 20%. The selected laser parameters for this study are summarised in the Table 3.

Analysis techniques
The surface morphologies of the samples were evaluated by mean of an Olympus BH2-UMA optical light microscope. The optical microscopy pictures were obtained at 20X magnification with DinoCapture 2.0 software.
The topography of the samples was analysed with the Bruker ContourGToptical profiler The assessing of the optical profilometry was carried out with green light, at 27.5X magnification. Figure 2 shows the images of the laser processed samples and mirror polished surface ( Fig. 2.a). The laser pulses at ≥ 130µJ created dimples with morphologies that were irregular circumference ( Fig. 2.b-d). This morphology can be due to the magnetic field of the material. The melting and evaporation of the material are usually the mechanisms that generate the dimples on metals for IR laser exposures [16]. These thermal processes are commonly featured by the ablation mechanism of hydrodynamic expansion for nanosecond pulse laser. The dimples of these ablation mechanisms are characterised by having melted material around the laser exposed area. The hydrodynamic expansion causes the melted material to mass on the dimple edges, which forms the circular shape of the dimple [43]. The melted material was accumulated both on the edges as well as the interior of the dimples (Fig. 2.b-d). This indicates that an external force had limited the hydrodynamic movements of the liquid material. With the exception of the gravity and magnetic field of the Earth, the magnetism of the Nd-Fe-B magnet was the only unique force applied on the liquid material. This force was responsible for the ejected material absence around the dimples and the irregular dimple forms because it retained the liquid material inside the laser exposed zone. This also indicated that the molten magnetic material is still being magnetically active [23,25].

Surface morphology
Some characteristics of the dimples were different according to E p . The dimples at ≥ 210µJ had ejected material outside shot area (Fig. 2c-d.). This can indicate that  these dimples are created by the ablation mechanism called as explosion phase. This mechanism is the nucleation and subsequent growth of a bubble that consists of evaporated material surrounded by liquid. The gas material causes pressure on the melted material. If this pressure is sufficiently elevated, the bubble can explode. This explosion ejects the molten material outside of the laser treated area [16,43,44]. The ejected material is absent on the dimples at 134µJ (Fig. 2.b.), which indicates that another mechanism produces these dimples, which is surface vaporisation. This ablation mechanism is featured by the transformation of the liquid to gas in a liquidgas interface that is often called the Knudsen layer. Although the escape velocity distribution of evaporated material in the liquid layer initially is non-equilibrium, this distribution becomes equilibrium with time [16,43]. The dissimilar ablation mechanism for dimples at 134µJ with respect to the dimples created at other E p is because the energy fluence necessary for each mechanism is different. Usually, the explosion phase requires more energy fluence than surface vaporisation [16]. The ejected material was so close to the dimple (Fig. 2c-d.) due to the magnetic force of the sample. This material can generally found within a certain distance of the laser hit area [45]. This distance is reduced by the magnetic force of the material because the molten material is magnetic active [23,25]. The Lorentz force can be opposite direction to the ejection force created by explosion phase. The distance between ejected material and the dimple was lower than expected distance. The width of the dimples (W), defined as the diameter of all the area treated by the laser pulse, was widened with the increment of E p , as can be in the graph of Fig. 3. This is because the energy spatial distribution is Gaussian like [26].
The focused laser beam diameter (d o ) and the pulse energy threshold for melting process (E th−melting ) were 53.8 ± 13.2 μm and 54.889 ± 0.004µJ, respectively. E th−melting represented the minimum laser energy necessary to melt the magnet. These values were estimated via Eq. 6 [26,28] and the liner regression method of the Fig. 3.b. graph data.
d o was slightly larger than to d o(Theo) (51 μm) because the ejected material increases W. Liquid material is ejected outside the pulse zone through explosion phase, which produces an increasing of the ablation rate [44]. This enhancement of the ablation rate means that the dimples are deeper and wider than in the case of the normal situation of evaporation. This causes that the growth rate of the W with the increasing of E p is greater than the theoretical value that was d o (Theo) . The experimental energy fluence threshold (φ th ) was 2.4 J/cm 2 calculated through Eq. 7.
This is due to the magnetic forces of the samples that can reduce W. As mentioned, the electromagnetic forces produced by solid material can attract the melted material to the centre of the laser impacted zone [25]. This produces a contraction of the melting material, which in turn narrows W. This is observed in the higher experimental energy fluence threshold than that that for theoretical case.

Topography
The topography of the dimples consists of a valley surrounded by a crest, as can be observed in Fig. 4. The valley is formed by the ablation of the material via the evaporation of metallic material and the movement of the molten material to outside laser shot area [18,24]. The crest around valley is created by mean of the accumulation of the liquid material on dimple rim and after its subsequent, fast, re-solidification. The pressure of the evaporated material on the molten displaces the liquid material to edge of the dimple [22,24]. The shape of the dimples is a spherical cap that is commonly called U-type [29][30][31][32]. This shape is due to only the Gaussian peak of the laser beam having sufficient energy to ablate the material [32].
The graphs of Fig. 5 show the diameter of the dimple (d) according to E p . d is defined as the diameter of the ablated area. Increasing E p also increased the d because of the Gaussian energy spatial distribution in laser beam [10,32]. d o (focused laser beam diameter) and E th (minimum laser energy to ablate the magnetic material) were estimated using Eq. 9 [10,28] while φ th (minimum laser energy fluence to ablate the magnet )was calculated using Eq. 7.
d o was found to be 41 μm, lower than d o(Theo) (51 μm), whose dissimilarity was 20%. This is because the magnetic forces of the samples that can diminish d o [23][24][25], as previously mentioned. In respect of φ th , its value was 3.756 J/cm 2 that was lower than φ th(Theo−evaporating) (13.190 J/cm 2 ). φ th(Theo−evaporating) was the minimum laser energy fluence to ablate the material, which was calculated through Eq. 10 [15,27,40]: The dissimilarity of the values is due to two reasons. The first reason is the reflectivity, R, of the mirror polished surface is different to the values of R from literature. The surface roughness can diminish the R of the samples because the laser beam can be scattered with > 180 degrees by peaks and valleys of the surface. Reflected laser radiations can impact again on the surface, which improves the absorption of the laser beam by material [14,28]. Secondly, there is a greater rate of ablation in the explosive phase than that for surface vaporisation. The explosion of the bubble increases the amount of the removed material due to the ejection of the molten material, which is characteristic of the explosion phase [44]. The explosion phase mechanism thus needs less energy than normal vaporisation to produce a similar ablation rate. φ th(Theo−evaporating) was calculated according to normal evaporation and therefore, this value is greater than experimental value.The increasing of E p deepened the cell, as can be seen in the graphs of Fig. 6. This is due to the amount of the ablated material being proportional to pulse energy [18]. Energy depth penetration (l) (the depth to which absorbed laser energy is transferred by thermal processes) and pulse energy threshold (E th ) (minimum pulse energy to produce a dimple with observable depth) for this case, were estimated using the graph of Fig. 6.b. and Eq. 11 [18,40], where D is the depth of the cell.
l and E th were 0.623 μm and 16.449µJ, respectively. l is the same as the theoretical energy depth penetration (l Theo ) for a geometric factor (ξ) of 1.3. This factor was calculated with Eq. 11 [17,26,27]. Figure 7 shows the graph of average roughness as function of the pulse energy. Mirror polished surfaces had an average roughness of 0.066 ± 0.003 μm. All surfaces with dimples were rougher than the original surfaces. The generation of a dimple on polished surfaces increases the relief, which rises the average roughness of the samples. The average roughness rose with the increment of the pulse energy because this feature of the samples is proportional to dimple size. Sample relief is proportional to dimple size and, it is in turn risen with the increment of E p . The rate of the ablation is greater at higher E p [18]. Thus, average roughness is proportional to E p .

Conclusions
The present work shows that the melted material of the NdFeB alloy is still magnetically active. Although this effect slightly impoverishes the quality of the textures, the dimples have an acceptable shape. The magnetic activity of the molten material indicates that this magnet can be designed by mean of other laser processes as, laser selective melting and laser additive manufacturing.
In addition, this study provides important information about the influence of the pulse energy on cell characteristics (W, d, D and average roughness), which was characterised through the determination of the laser beam-material interaction factors (l, E th and φ th ). The quantification of these factors can allow the design of features of the textures generated by single pulses on NdFeB alloy. This can therefore allow the Fig. 7 Graphs of the average roughness as function of pulse energy Fig. 6 Graphs of the pulse energy influence on depth, (a) pulse energy vs. depth and (b) pulse energy natural logarithm vs. depth creation of functionalised NdFeB alloy surfaces with desired properties, which can have numerous applications, e.g. in manufacturing and electronic engineering.
Author contributions All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by Juan Ignacio Ahuir-Torres. The first draft of the manuscript was written by Juan Ignacio Ahuir-Torres and Juan Ignacio Ahuir-Torres, and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.
Funding The authors declare that no funds, grants, or other support were received during the preparation of this manuscript.

Data Availability
The datasets generated during and/or analysed during the current study are available in the manuscript.