Analysis of anti-reflection mechanisms of the black aluminum alloy made by femtosecond laser processing

Mitigating the optical reflection of aluminum alloy over a broad spectral range from 0.45 μm to 15 μm is vital for many applications. This can be realized by introducing efficient light-absorbing textured surfaces via femtosecond laser surface processing. However, a clear analysis of antireflection performance has not been reported yet. This paper proposes a numerical model of anti-reflective structures is proposed based on SEM and EDS characterization. Multiple anti-reflective mechanisms were revealed intuitively through FDTD simulation.


Introduction
Mitigating metals' optical reflection is vital for many applications: stray light suppression for optical satellites [1] and stealth [2]. Conventional antireflection strategies such as emulsion, electrolytic oxidation, and ink print suffer from limited bandwidth and instability. Integrating wideband metamaterial absorbers on aluminum alloy suffers from high cost and low yield [3]. Vantablack, an ultra-black coating made by chemically growing carbon nanotubes, is too fragile for many practical applications [4].
By fabricating micro-nano hierarchical structures(HSs) on metal surfaces via direct femtosecond laser irradiation, black metals acquire the capability to suppress reflection over a broad spectral band [5][6][7][8][9][10][11][12]. Black metals have the advantage of long-term stability and structural robustness for practical application [13]. Aluminum alloy is commonly used in space-borne platforms, and laser processing toward highly anti-reflective black aluminum alloy is crucial for aerospace applications [14,15]. However, the aluminum alloy is highly reflective, and the low intrinsic loss makes it challenging to fabricate efficient light-absorbing textured surfaces on aluminum alloy [16,17]. A detailed investigation into the antireflection mechanism of the black aluminum alloy is crucial for guiding the design of efficient absorbing structures and the selection of laser processing parameters.
Formation mechanisms of laser-induced periodic surface structures(LIPSS) [17][18][19][20] and hierarchical structures [8,11,21] on metals are extensively studied, antireflection mechanisms and relevant simulated models were proposed. Tetsuo et al numerically investigated the light-absorbing nanostructures using finitedifference-time-domain (FDTD) [22] and claimed that the absorption of nanosecond processed copper is attributed to multi-scattering of the aggregated nano-particles [23]. Fan et al proposed the nano-wire-assisted phonon absorption and calculated the reflectance using the effective index method [24]. Andrew et al proposed a basic simulation model of surface structures on laser processed aluminum and calculated the effect of oxidation layer thickness on absorptivity using COMSOL [25]. Xin et al modeled LIPSS on the black stainless steel and assessed the reflection properties using the FDTD method, claiming that the absorption results from the resonance in wavelength-scale periodic surface structures and plasmonic absorption at subwavelength Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI. nano-particles [26]. It's well known that broadband absorption of hyper-hierarchical structures is attributed to the light trapping effect [27] of multi-scale structures [5]. However, those modeling and analysis of anti-reflective structures are incomplete and qualitative. This paper analyzes quantitively the broadband antireflection mechanisms of the black aluminum alloy made by femtosecond laser processing. Firstly, morphologies and element composition of the fabricated antireflective structures are characterized using scanning electron microscopy (SEM), Energy-dispersive x-ray spectroscopy (EDS), and focused ion beam (FIB). Secondly, a numerical model was proposed to simulate the spectral reflectance of the anti-reflective structures, and multiple antireflection mechanisms were revealed through the electric-field distribution in the HSs.

Fabrication, characterization, and modeling
The center wavelength of the femtosecond laser (Femto-IR-80-60) is 1035 nm, and the femtosecond laser is linearly polarized. Anti-reflective structures were fabricated using an x-y galvanometer and F-θ lens (TSL-1064-10-56Q-D20, the effective focal length is 56 mm, and the diameter of the focus spot is 9 μm). A polished 2A12 aluminum alloy plate with a thickness of 4 mm was cleaned using acetone before laser processing. The aluminum alloy plate was irradiated in two-dimension crossing scans by the femtosecond laser with processing parameters shown in table 1. Anti-reflective structures were generated on the aluminum alloy plate via two-step femtosecond laser irradiation in the air [12]. Regular macro moth-eye structures were created during the multiple fast laser scanning process, and abundant nano-particles were introduced on the surface during the single slow scan [24, 28].

Characterization
As shown in figures 1(a), (b), the morphologies of the anti-reflective structures are parabolic-like, with a period of about 13-14 μm. The femtosecond laser is linearly polarized, and LIPSS is not noticeable on the top surface, as shown in figure S5. The SEM image of the surface structures and the compositional distribution of carbon, oxygen, and aluminum across surface structures are shown in figure S2. As shown in figure S2(b), the carbon content is close to zero, and the primary element of surface structures are oxygen and aluminum, as shown in figure S2(e).
The anti-reflective structures' internal content was investigated using FIB (FEI Helios NanoLab G3) crosssectional cutting and EDS analysis [29]. A protective platinum layer with a thickness of 500 nm is deposited to protect the underlying materials from the focused Ga + ion beam milling. Two milling steps are performed involving milling at a high current (9.3 nA) followed by polishing at a low current (0.23 nA). SEM image of crosssectioned structures shows that the anti-reflective structures are composed of bulk aluminum alloy and the  redeposited layer [25], and the thickness of the redeposited layer is about 3 μm, as shown in figure 1(c). It's reported that deep moth-eye structures were formed on the LIPSS-covered surface, and the ejected nanoparticles deposited back on the surface, which buried the LIPSS pattern [30]. The roughnesses on the surface are micro-scale and nano-scale protrusions formed during redeposition instead of LIPSS [21].
In figure 2(a), the cross-section of another surface structure labeled by the dashed line clearly shows the boundary between the bulk aluminum structure and the redeposited layer. The EDS analysis of the redeposited layer labeled by the yellow dashed line in figure 4(b) shows that the redeposited layer is a mixture of Al 2 O 3 and aluminum. Figure 3 illustrates how the redeposited layer is grown on bulk aluminum moth-eye structures during multiple fast laser scanning processes [30,31]. The bulk aluminum moth-eye structures were formed by introducing surrounding grooves via laser ablation [6,18,32]. Aluminum alloy nano-particles generated during laser ablation deposited back onto the surface as the laser beam moved away, and new nano-particles were generated as the laser beam returned, leading to the redeposited layer [33,34]. The generated nano-particles partially oxidized, and the redeposited layer is a mixture of Al 2 O 3 and aluminum.   line). The aluminum moth-eye structure is assumed to be parabolic in shape with the height of 'H', covered by the Al 2 O 3 layer with the thickness of 't p ' at the top, and 't s ' on each side and the diameter of micro-particle is 'D 1 '.

Modeling
(2) Roughnesses on the surface were treated as micro-nano tree-like protrusions [27], as labeled by the red dotted line in figure 4(a). Micro-scale protrusions are distributed uniformly on the moth-eye structure with a horizontal and vertical distance of 'G'. As shown in figure S4, the major elements of micro-scale protrusion are oxygen and aluminum. For simplicity, the micro-scale protrusion is assumed to be composed of aluminum alloy hemispheroid with the diameter of 'D 1 ', covered by the Al 2 O 3 layer with the of 't 1 '. Similarly, nano-scale protrusions distributed on the micro-scale protrusion are composed of aluminum alloy hemispheroids with a diameter of 'D 2 '.
(3) As shown in figure S3, deep holes between adjacent hierarchical structures are simplified into cylinder holes decorated with nano-particles. Figure 4(b) shows the configuration of the simulation, the hierarchical structures are assumed to be periodically arranged in a square lattice, and such a periodic array is studied by simulating a unit cell of the array with periodic boundary conditions applied at the x-and y-boundaries and perfectly matched layers (PMLs) applied at the z-boundaries. A plane wave source linearly polarized in the x-direction is used to excite the structure, and a power monitor is used to collect the reflected waves. The mesh size is set to 60 nm, and the simulated reflectance does not change significantly using finer mesh (40 nm).
We calculated the average reflectance of seven hierarchical models with different parameters (table 2) and compared it with measured reflectance. Parameters are adopted based on the SEM characterization: (1) The laser scanning interval is 13 μm, and the period of fabricated hierarchical structures is about 13-14 μm, as shown in figure S1.    (2) As shown in figure 1(c), the thickness of the redeposited layer is about 3 μm, 't p ' is selected within 2.5-3.5 μm, and 't s ' is selected within 1.5-2 μm.
(3) As shown in figure S3, the height of hierarchical structures is about 26 μm, and 'H' is selected within 19-22 μm.
(6) As shown in figure S6, the distance between the valley of the hole and the peak of moth-eye structures is about 30 μm, and 'D' is set to be 2 μm.

(7)
The uncontrollable value of 't 1 ' and 'G' is selected empirically so that the simulated reflectance agrees with the measured reflectance.

Simulation and discussion
The total hemispherical reflectance (THR) of black aluminum alloy over 0.45-2.5 μm is characterized by the SolidSpec-3700 spectrophotometer from Shimadzu Corp. with an integrating sphere. We measured the spectral reflectance over 2.5-15 μm using Nicolet iS50R (from Thermo Scientific) with an integrating sphere. The black  aluminum alloy shows comprehensive antireflection performance with measured reflectance of 5.9% and 9.1% on average over the 0.45-2.5 μm and 2.5-15 μm bands. As shown in figure 5(b), the simulated reflectance fluctuates around the measured reflectance over 2.5-15 μm, which verifies the validity of the simulated model. The simulated reflectance may fit better with the measured when more simulated models are considered.
As shown in figure 5(a), the simulated reflectance is 1.25% greater than the measured reflectance on average over 0.4-2.5 μm because fine nano-particles (diameter below 200 nm) cannot be modeled due to finite computational resources. It's proved that fine nano-particles make a great contribution to anti-reflection performance over VIS-NIR.
To further reveal the anti-reflection mechanisms, the local distribution of electric field at 14 μm, 8 μm, 5 μm and normalized energy loss density at 0.8 μm were plotted in figure 6.
(1) As shown in figure 6(a), the multiple horizontal and uniform bright spots are the proof of the interference between the reflected light and the incident light, indicating that the incident photon at 14 μm was trapped in macro-cavities formed in the inter-moth-eye areas [35]. The stronger electric field at the sidewalls infers that light rays bounce more times between the sidewall of the cavities [5, 6].     As shown by figure S9, phonon-assisted absorption in the Al 2 O 3 layer makes the sidewall much more absorptive, suppressing reflectance over 10-15 μm [24, 36].
(3) As shown in figures 6(c) and S8, the stronger electric field at the nano-particles indicates that nanoparticles increase reflection times through the scattering mechanism and enhance the light trapping effect over 3-5 μm.
Micro-nano cavities in the intra-moth-eye area act as porous absorbers [39] and significantly suppress the reflection over Mid-infrared [35,40].
(4) As shown in figure 6(d), photons at 0.8 μm was absorbed on the surface of Aluminum nano-particles, which is attributed to surface plasmons. The aggregates of nano-particles with different diameters lead to broadband absorption over VIS-NIR [41,42].

Further discussion
It's reported that macro-parameters (period, height, thickness of oxidation layer) and micro-parameters (the diameter and abundance of micro-particles) can be controlled by laser processing parameters. In this section, we examine the influence of macro-micro-parameters on the antireflection performance, which will provide guidance to improve its anti-reflection performance. As shown in figure 7(a) and table 3, simulated reflectance over 2.5-15 μm decreases as the period reduces to 11 μm (cyan), and a smaller period may be preferred for better anti-reflection performance. As shown in figure 7(b), hierarchical structures with a period of 11 μm have a stronger light-trapping effect over 10-15 μm, indicated by the stronger electric field. The stronger light-trapping effect maximizes the phonon-assisted absorption and suppresses the reflectance to a minimum. Meanwhile, hierarchical structures with a period of 11 μm and sharper tips have lower reflection at the top of hierarchical structures (red dotted line in figure 7(c)).
It's reported that the thickness of the oxidation layer in the redeposited layer can be controlled by the oxygen content of the atmosphere where the sample was processed. As shown in figure 8 and table 4, reflection over 10-15 μm decreases rapidly as the oxidation layer thickens, which is attributed to the stronger phonon-assisted absorption. The thicker oxidation layer and lower reflectance over 10-15 μm can be predicted when the sample is processed in the oxygen-rich atmosphere with increased fluence.
As shown in figure 9(a) and table 5, the anti-reflectance performance maintains over 9-15 μm and degrades over 3-9 μm. As indicated by clear bright spots in figure 9(b), the scattering effect weakens, because the distance between nano-particles on the surface of micro-particles increases as D 1 increases.
As shown in figure 10(a) and table 6, adjusted Model-I with micro-particles sparsely distributed (G = 1.5D 1 ) has better anti-reflection performance over 10-15 μm because of a larger photon-Al 2 O 3 contact area and stronger light-phonon interaction. Degraded anti-reflection performance over 5-10 μm is attributed to lowabundance micro-nano particles and weakened scattering effect, inferred by regular and horizontal bright spots in figure 10(b).
Femtosecond laser processing with a smaller scanning interval and fluence is preferred to introduce hierarchical structures with a smaller period and a thicker oxidation layer, which have better anti-reflection performance. Micro-nano-particles with an appropriate diameter and density are essential for overall antireflection performance, and precise control of micro-nano parameters will be fruitfully explored in future studies.

Conclusion
This paper proposes a complete numerical model to investigate the broadband anti-reflective performance of black aluminum alloy through morphologies measurement, EDS analysis, and internal content analysis. Multiple absorption mechanisms were revealed using the FDTD method intuitively. This paper provides a new approach to understanding the antireflection mechanisms qualitatively of black metals and designing surface structures for special demands.