Introduction

The global transition to renewable energy has positioned silicon-based solar modules as a cornerstone technology. However, the long-term efficiency and yield of such systems are significantly constrained by two major surface phenomena: front-glass reflection and soiling. Studies show that uncoated cover glass typically loses about 4% of incident sunlight due to reflection at the air–glass interface1,2,3,4.

Conversely, soiling, defined as the accumulation of dust, dirt, and other particulates on the module surface, can cause annual energy losses of approximately 1.5 to 6% in well-maintained systems, while losses greater than 15% have been documented under severe arid or desert environments5,6,7,8. These losses not only reduce energy production but also increase cleaning and maintenance demands, compromising sustainability and economics. Accordingly, there is a clear need for durable, multifunctional surface coatings for PV modules that combine anti-reflective (AR) behavior with self-cleaning or anti-soiling functionality.

Titanium dioxide (TiO2) is widely regarded as a benchmark photocatalyst for self-cleaning surfaces due to its strong oxidative capability, high stability, and low cost9,10. Nevertheless, TiO2 is not an ideal material for spark-based deposition. Its high heat of fusion and large specific heat capacity requires substantially more thermal energy for melting and evaporation compared with tin. As a result, TiO2 nanoparticles form slowly during spark discharge, leading to a significantly reduced deposition rate and lower areal coverage per spark cycle. These limitations hinder its scalability for large-area fabrication using spark ablation, particularly when industrial throughput is required. This practical constraint motivates the search for alternative metal oxides that offer both strong photocatalytic functionality and superior compatibility with spark-based synthesis.

Recent studies show that SnO2-based thin films can serve as effective anti-reflection layers while also providing self-cleaning or antifogging functionality. For example, Chandralekha et al. (2024) reported spin-coated SnO2 films modified with silane compounds achieving ~ 92% transmittance and enhanced dust-resistance11. Even patent literature confirms the viability of SnO2 nanoparticle coatings for anti-reflection solar glass with enhanced durability (US10059622B2)12.

In addition to its optical advantages, tin exhibits distinctly more favorable thermal properties for spark-based nanoparticle generation. Compared with titanium, metallic tin has a much lower heat of fusion (7.03 kJ/mol vs. 14.15 kJ/mol) and a significantly smaller specific heat capacity (≈ 0.23 J/g K vs. ≈0.52 J/g K), enabling far faster melting and vaporization during each discharge event13. As a result, SnOx nanoparticles form more readily and at much higher deposition rates, making Sn-based coatings more compatible with large-area, high-throughput spark ablation processes.

Traditional deposition methods for SnO2-based films-such as sputtering14, CVD15, and spray pyrolysis16 often require high temperatures, vacuum chamber or chemical solvents. In contrast, the sparking process offers a compelling, single-step, and green alternative, enabling rapid synthesis of high-purity, nanocrystalline metal oxide films at ambient conditions without chemical precursors. This method is inherently scalable and compatible with large-area substrates, addressing a key industrial requirement17,18,19.

In practical photovoltaic operation, modules are continuously exposed to airborne organic contaminants, environmental pollutants, and dust-bound hydrocarbons that adhere to the glass surface and reduce light transmission. While surface wettability assists in the physical removal of particulates, photocatalytic functionality plays a critical role in decomposing organic residues under solar illumination. The in situ generation of reactive species enables gradual breakdown of fouling layers, facilitating removal by rain or water flow and contributing to sustained power conversion efficiency during long-term outdoor exposure.

Building upon our previous work on titanium-based spark coatings4, we found that replacing Ti with Sn significantly accelerates deposition-achieving a threefold increase in coating speed while maintaining multifunctional surface properties. Incorporating an external magnetic field further enhances nanoparticle transport toward the substrate, yielding an additional 1.75-fold increase in deposition rate. Overall, this strategy provides a 5.25× improvement in coating throughput, highlighting the scalability potential of magnetic-field-assisted spark ablation.

In this work, the external magnetic field is employed not only to accelerate deposition but also as a control parameter for phase and morphology engineering during spark ablation. By aligning the magnetic field parallel to the substrate surface, uniform magnetic-field-assisted coating over full-size photovoltaic glass is achieved. We hypothesize that magnetic confinement modifies plasma plume dynamics, influencing nanoparticle cooling and oxidation behavior, and thereby tuning the SnO2–SnO phase ratio and aggregation characteristics. Structural analyses (XRD, Raman, and HRTEM) confirm the formation of a biphasic SnO2–SnO composite under magnetic assistance. The resulting films are systematically evaluated for optical, self-cleaning, and anti-soiling performance, demonstrating an optimized multifunctional coating strategy for large-scale photovoltaic applications. Photocatalytic functionality is particularly important because a significant fraction of soiling on photovoltaic modules contains organic components, including airborne hydrocarbons, pollen residues, and combustion-derived deposits, which cannot be fully removed by simple water rinsing. Under solar illumination, photocatalytic oxide surfaces generate reactive oxygen species that oxidatively decompose these organic contaminants into smaller, less adhesive species, thereby facilitating their subsequent removal by rain or surface water flow and contributing to sustained long-term optical transparency.

Experimental

Film preparation

The home-built sparking system used for nanoparticle film deposition is illustrated in Fig. 1. The apparatus was equipped with a four-pair spark-tip assembly designed for a 1 × 1 cm2 quartz substrate. The anode–cathode gap was fixed at 1 mm. During operation, the spark head was scanned in two dimensions at a controlled velocity (typically 40–70 mm min-1 unless otherwise specified), while the substrate was positioned 3 mm below the electrodes in ambient air. Titanium wire (Ø 0.5 mm, 99.8% purity) and tin wire (Ø 1.0 mm, 99.95% purity), purchased from Advent Research Materials Ltd., were used as the sparking electrodes. A DC voltage of approximately 5 kV, discharged from a 24 nF capacitor, generated the spark plasma.

Fig. 1
Fig. 1The alternative text for this image may have been generated using AI.
Full size image

Schematic diagram of the spark-coating system used for the deposition of Sn and Ti nanoparticle films.

For magnetic-field-assisted deposition, permanent magnets were arranged beneath the substrate as described in Fig. 2. The magnetic-field strength at the substrate surface was controlled by adjusting the magnet–substrate spacing, yielding field intensities of 220–350 mT depending on the experimental condition. Unless otherwise stated, a 300 mT field was used for magnetic optimization studies, while 220 mT was selected for durability evaluation to minimize thermal stress during high-speed coating.

Fig. 2
Fig. 2The alternative text for this image may have been generated using AI.
Full size image

(a) Schematic configuration of the magnetic-field-assisted spark system showing electrode arrangement, magnet placement, and electrode–substrate spacing. (b) Lateral magnetic-field distribution at different electrode–substrate gap distances.

Muddy water preparation and muddy testing

Muddy water was prepared by dispersing fine soil particles in water to achieve a concentration of 10 g/L20. Three coated and uncoated solar panels were used in the experiment. Each panel was fixed at the optimal year-round tilt angle of 19° for Chiang Mai, Thailand21. One liter of muddy water was sprayed onto each panel using a hand-held sprayer. After drying under ambient conditions for 2 h, muddy-water testing was performed.

Characterizations

A home-made tensiometer was used to measure the water contact angle (WCA). Surface morphology and film thickness were examined using scanning electron microscopy (SEM, JEOL JSM-6335 F). The crystalline structure of the nanoparticle films was analyzed by X-ray diffraction (XRD, Rigaku SmartLab, Japan). Raman spectroscopy (JOBIN YVON HORIBA T6400) with a 514 nm Ar-ion laser operated at 50 mW was used to determine the phase composition. Optical properties were measured over a wavelength range of 250–800 nm using a UV–vis spectrophotometer (Hitachi U-4100). Photocatalytic activity was evaluated by monitoring the degradation of 0.01 mM methylene blue (MB) solution after exposure to natural sunlight for 5 h. The power conversion efficiency (PCE) of the solar panels was measured using a solar module analyzer (PROVA 210) equipped with an array of 860 W halogen lamps.

Results and discussion

Figure 3 illustrates the evolution of water contact angles (WCAs) on the spark-generated Sn and Ti oxide films as a function of repeated scanning cycles at a scanning speed of 5.25 mm/s. Both materials exhibit a monotonic decrease in WCA with increasing deposition cycles, reflecting the progressive buildup of porous oxide networks that facilitate capillary-driven wetting. However, the transition to superhydrophilicity occurs at markedly different rates. For the Sn coating, the WCA rapidly drops from ~ 50° to nearly 0° after a single scanning cycle, indicating the formation of a sufficiently porous and interconnected SnOx layer in just one pass. In contrast, the Ti coating requires approximately three cycles to reach comparable superhydrophilic behavior. This disparity arises from fundamental thermal-property differences between the two metals: tin possesses a significantly lower heat of fusion and specific heat capacity than titanium, enabling it to melt, vaporize, and generate oxide nanoparticles more efficiently during each spark discharge. As a result, Sn nanoparticles accumulate and coalesce more rapidly, forming a dense oxide layer that enhances capillary wicking at a much earlier stage. These results confirm that Sn is inherently more suitable for fast spark-based coating processes, allowing the production of superhydrophilic surfaces with minimal deposition cycles, while Ti requires substantially more processing time to achieve the same wetting performance. The minimum number of deposition cycles required to achieve superhydrophilicity for each material - one cycle for Sn and three cycles for Ti - was therefore selected as the processing condition for all subsequent evaluations of surface morphology, optical properties, and photocatalytic activity.

Fig. 3
Fig. 3The alternative text for this image may have been generated using AI.
Full size image

Variation of water contact angle (WCA) on Sn- and Ti-coated surfaces as a function of repeated scanning cycles at a scanning speed of 5.25 mm s-1.

The SEM images in Fig. 4a,b show the surface morphology of spark-generated Sn and Ti coatings deposited at a scanning velocity of 40 mm min-1 without magnetic assistance. Under identical processing conditions per pass, the Sn film achieves comparable surface coverage in a single scanning pass, whereas the Ti film requires three sequential passes. Both coatings exhibit porous nanoparticle-based structures; however, the Sn layer consists of larger and more loosely packed aggregates, consistent with its more efficient melting and nanoparticle generation during each spark discharge. In contrast, the Ti coating displays smaller and more uniformly distributed nanoparticles, reflecting its slower material consumption and nanoparticle production rate. The cross-sectional SEM images in Fig. 4c,d further demonstrate that the Sn coating forms a thicker porous layer per pass, while the Ti film remains comparatively thinner under the same single-pass condition, thereby necessitating additional passes to achieve similar coverage. Figure 4e–h present the corresponding morphologies under magnetic-field assistance (300 mT), where enhanced nanoparticle accumulation and more pronounced film buildup are observed. Overall, these results indicate that the intrinsic thermal properties of tin promote more rapid film formation per scanning cycle, supporting the improved coating throughput discussed later. The magnetic-field configuration used for the assisted deposition is illustrated in Fig. 2.

Fig. 4
Fig. 4The alternative text for this image may have been generated using AI.
Full size image

SEM characterization of spark-generated Sn and Ti coatings deposited at a scanning velocity of 40 mm min-1 under non-magnetic and magnetic-field-assisted conditions. (a,b) Top-view images of Sn (1 pass) and Ti (3 passes) films deposited without magnetic field. (c,d) Corresponding cross-sectional SEM images showing film thickness and porous oxide-layer formation. (e,f) Top-view images of Sn and Ti films deposited under a 300 mT magnetic field (1 pass). (g,h) Corresponding cross-sectional SEM images of magnetically assisted coatings.

From an industrial perspective, coating throughput is a critical factor governing scalability22. Sparked tin offers a clear practical advantage: the deposition time required to produce a functional SnO2 film is significantly shorter than that of TiO2. This difference arises from the lower heat capacity and heat of fusion of Sn compared with Ti, which enable more efficient melting, vaporization, and nanoparticle generation during each spark event. Although TiO2 is widely recognized for its photocatalytic performance, the present results demonstrate that spark-generated SnO2 films achieve comparable photocatalytic activity while providing superior processing efficiency. Considering both functional performance and manufacturing practicality, Sn was selected for further optimization.

To further increase coating throughput on large-area substrates, a magnetic-field-assisted configuration was developed. While previous studies have shown that external magnetic fields can enhance nanoparticle deposition23, those approaches were limited to small substrates placed directly on permanent magnets. Such configurations are unsuitable for commercial solar panels, as magnets positioned above and below a thick substrate would both weaken the effective field at the spark–substrate interaction zone and obstruct lateral scanning of the multi-pin spark head.

To overcome these limitations, a redesigned magnetic configuration was implemented (Fig. 2a), in which permanent magnets are arranged so that the magnetic field lines run parallel to the spark electrodes. The solar panel is positioned parallel to the magnet surface and separated by a thin PTFE sheet, allowing free lateral motion beneath the electrodes while maintaining a controlled magnetic environment. By adjusting the magnet–substrate spacing to 1.1, 2.2, and 3.1 mm, magnetic-field strengths of 350, 300, and 250 mT were achieved at the substrate surface, respectively (Fig. 2b). The field remains laterally uniform over approximately ± 25 mm from the spark-head center, ensuring stable coating conditions during scanning.

Figure 5 shows the variation of water contact angle (WCA) as a function of scanning velocity under different magnetic-field strengths. For a given scanning velocity, increasing magnetic-field intensity reduces the WCA, indicating enhanced nanoparticle accumulation and improved film formation at higher coating speeds. This trend is consistent with Lorentz-force-assisted plasma confinement, in which charged species are more effectively guided toward the substrate under stronger magnetic fields. Consequently, superhydrophilicity (WCA ≈ 0°) is maintained over a broader range of scanning velocities, enabling higher-throughput processing suitable for industrial-scale fabrication.

Fig. 5
Fig. 5The alternative text for this image may have been generated using AI.
Full size image

Water contact angles of SnOx films deposited under different magnetic-field strengths as a function of spark-head scanning velocity.

However, the sample deposited at 350 mT exhibits inferior performance compared with those prepared at 200–300 mT. Although stronger magnetic confinement is expected to enhance deposition efficiency, achieving 350 mT required reducing the magnet–substrate gap to approximately 1.1 mm, resulting in an extremely short spark–substrate distance (~ 1 mm). Under this condition, thermal energy from the plasma and molten tin droplets is transferred to the substrate with limited dissipation, likely causing localized overheating, surface damage, or microstructural irregularities that disrupt uniform nanoparticle assembly.

Additionally, excessive magnetic confinement at such a short separation may alter plasma dynamics, potentially leading to premature plume collapse, reduced oxidation volume, and modified nanoparticle trajectories. The combined effects of thermal stress and altered plasma behavior likely account for the degraded wettability observed at 350 mT. These results indicate that, although magnetic fields enhance coating efficiency, a practical upper limit exists due to geometric and thermal constraints. In the present system, 300 mT provides the optimal balance between enhanced deposition and substrate integrity, and this condition was therefore selected for subsequent experiments.

Photocatalytic degradation of methylene blue (MB) under natural sunlight in Chiang Mai (8:00–13:00, 5 h) was evaluated for SnO2 and TiO2 coatings with and without magnetic-field assistance (300 mT), as shown in Fig. 6a. The characteristic MB absorption peak at 665 nm decreases progressively with irradiation time for all coated samples, whereas the MB control exhibits only minimal change, confirming the catalytic role of the oxide films. The absorbance reduction for SnO2 and TiO2 is comparable under identical conditions. The corresponding relative concentration profiles (C/C0) in Fig. 6b further confirm this trend. Both SnO2 and TiO2 coatings show a continuous decline in MB concentration over the 5-hour exposure period, with closely overlapping degradation curves. The magnetically assisted samples exhibit similar photocatalytic behavior, indicating that spark-generated SnO2 maintains photocatalytic performance comparable to benchmark TiO2 under the same illumination conditions.

Fig. 6
Fig. 6The alternative text for this image may have been generated using AI.
Full size image

Photocatalytic degradation of methylene blue (MB) under natural sunlight using spark-generated SnO2 and TiO2 coatings with and without magnetic-field assistance (300 mT). (a) UV–vis absorbance spectra showing the decrease of the characteristic MB peak at 665 nm. (b) Corresponding relative concentration curves (C/C0) as a function of irradiation time, including MB control and dark control.

Applying a 300 mT magnetic field during spark generation modifies the nanoscale architecture of the SnOx products. The XRD patterns (Fig. 7a) confirm that tetragonal SnO2 (JCPDS 41-1445) remains the dominant crystalline phase in all samples, with characteristic reflections at the (110), (101), and (211) planes. Both conditions exhibit broad diffraction peaks consistent with nanocrystalline domains. Although the magnetically assisted sample shows a relatively stronger (110) reflection, absolute peak intensity can be influenced by factors such as nanoparticle coverage and film thickness. Therefore, this difference is not solely attributed to enhanced crystallinity, and phase identification is primarily supported by complementary Raman and HRTEM analyses (Fig. 7b–f).

Fig. 7
Fig. 7The alternative text for this image may have been generated using AI.
Full size image

Structural characterization of spark-generated SnOx nanoparticles synthesized without and with magnetic-field assistance: (a) XRD patterns, (b) Raman spectra, (c,d) TEM images, and (e,f) HRTEM images showing phase distribution and nanoscale morphology.

The corresponding Raman spectra (Fig. 7b) provide further evidence of phase composition. The non-magnetic sample exhibits dominant vibrational modes characteristic of SnO2, particularly the Eg mode near ~ 565 cm-1. In contrast, the magnetically assisted sample shows additional features associated with SnO, including characteristic modes in the ~ 110–220 cm-1 range, confirming the coexistence of SnO2 and SnO phases. The relative enhancement of SnO-related peaks under magnetic assistance is consistent with the formation of a mixed-valence SnO2–SnO heterostructure, in agreement with the HRTEM observations.

TEM observations (Fig. 7c,d) reveal a clear morphological contrast between the two samples. Both were directly deposited onto TEM carbon-coated copper grids under identical conditions, without solvent dispersion or post-treatment, ensuring that the observed aggregation differences arise from the spark-generation process rather than preparation artifacts. Nanoparticles generated without magnetic assistance remain relatively dispersed, whereas those synthesized under magnetic confinement form dense and interconnected aggregates. This aggregation behavior is consistent with enhanced plasma confinement, which increases particle–particle interactions and promotes partial sintering during deposition.

HRTEM images (Fig. 7e,f) further reveal differences in phase distribution. The non-magnetic sample predominantly exhibits lattice fringes corresponding to SnO2 (110), whereas the magnetically assisted sample shows intimate coexistence of SnO2 (110) and SnO (101/111) domains, forming a biphasic nanostructure. This observation suggests that magnetic confinement influences oxidation dynamics during the flight and cooling of molten Sn droplets in ambient air. The increased plasma residence time and modified solidification kinetics likely facilitate partial stabilization of Sn2+ alongside Sn4+, resulting in a mixed-valence SnO2–SnO heterostructure rather than a predominantly single-phase oxide.

The transmittance spectra of the superhydrophilic SnOx nanoparticle films deposited under magnetic assistance on quartz substrates are presented in Fig. 8a. The maximum transmittance in the visible range increases from 91.8% for the uncoated substrate to approximately 93% after coating. This improvement indicates that the ultrathin SnOx nanoparticle layer does not behave as a conventional absorbing film. Instead, it introduces a mild antireflective effect that suppresses Fresnel reflection at the air–quartz interface, thereby enhancing overall optical transmission24. The reflectance spectra (Fig. 8b) further support this observation. The coated sample exhibits a reduced maximum reflectance of approximately 8.5%, compared with 9.0% for the bare substrate. This reduction is likely associated with nanoscale surface texturing and a gradual refractive-index transition introduced by the deposited nanoparticle layer. Such morphological features can decrease Fresnel reflection at the interface while maintaining high transparency. Overall, the optical measurements confirm that the SnOx coating slightly suppresses surface reflection without compromising visible-light transmission. This modest antireflective contribution is consistent with the observed 4.02% enhancement in power conversion efficiency (PCE) under clean conditions.

Fig. 8
Fig. 8The alternative text for this image may have been generated using AI.
Full size image

Optical properties of SnOx-coated quartz substrates: (a) transmittance spectra and (b) reflectance spectra compared with the uncoated substrate.

The I–V characteristics of the coated and uncoated solar panels are presented in Fig. 9(a). The uncoated panel delivered a maximum power output of 5.23 W, whereas the coated panel reached 5.45 W, corresponding to a 4.02% increase in power conversion efficiency (PCE) under clean conditions. This enhancement is consistent with the modest reduction in surface reflectance and improved light transmission introduced by the SnOx nanoparticle coating. The self-cleaning behavior under muddy-water exposure is shown in Fig. 9(b). The uncoated panel retained a substantial amount of soil residue, resulting in widespread surface staining. In contrast, the coated panel exhibited rapid droplet spreading and drainage, leaving the surface visibly cleaner. The drying time was significantly reduced, with the coated panel drying within 15 min compared to nearly one hour for the uncoated panel. The impact of soiling on device performance is summarized in Fig. 9(c). After muddy-water spraying and natural drying, the maximum power of the coated panel decreased by 2.57% (from 5.45 W to 5.31 W), whereas the uncoated panel exhibited a larger reduction of 4.21% (from 5.23 W to 5.01 W). When expressed relative to the uncoated condition, the coating provides a 6.00% performance advantage after mud exposure, demonstrating its ability to mitigate soiling-induced efficiency losses.

Fig. 9
Fig. 9The alternative text for this image may have been generated using AI.
Full size image

(a) I–V characteristics of uncoated and SnOx-coated solar panels. (b) Photographs of coated and uncoated panels after muddy-water spraying. (c) Maximum output power before and after muddy-water exposure.

The enhanced photovoltaic self-cleaning and anti-soiling behavior arises from the synergistic interaction between photocatalytic activity and superhydrophilic surface properties. The biphasic SnO2–SnO architecture facilitates interfacial charge separation, promoting reactive oxygen species generation under solar illumination and accelerating degradation of organic contaminants, as supported by the methylene blue degradation results (Fig. 6). Simultaneously, the magnetically assisted films exhibit persistent superhydrophilicity (WCA ≈ 0°, Fig. 6), enabling rapid water spreading and efficient removal of soil residues (Fig. 9b). Although surface hydroxyl density was not directly quantified, the observed wettability and contaminant removal behavior indicate enhanced surface adsorption activity. The improved self-cleaning and anti-soiling behavior can therefore be understood as a coupled effect: magnetic-field-assisted deposition modifies the SnO2/SnO phase distribution and aggregation, which may promote interfacial charge separation and reactive oxygen species generation under illumination, while the superhydrophilic surface ensures rapid water spreading and contaminant removal. Together, these effects sustain optical transparency and mitigate soiling-induced efficiency losses.

After deposition under a 220 mT magnetic field at a scanning speed of 60 mm min-1, the Sn coating exhibits a dense and highly interconnected nanoparticle (NP) network uniformly distributed across the glass substrate (Fig. 10a). This magnetic-field strength was selected for durability evaluation to balance enhanced deposition efficiency with reduced thermal stress on the substrate, as discussed earlier. Following 1,000 droplet impacts from a height of 30 cm, only slight removal of weakly adhered surface particles is observed, while the underlying NP network remains largely preserved (Fig. 10b). Even after 5000 and 10,000 impacts (Fig. 10c,d), only minor morphological rearrangements occur, with most Sn nanoparticles remaining firmly bonded to the substrate. The coating thus maintains continuous nanoscale coverage after severe liquid impact, demonstrating strong mechanical robustness. These results indicate that the spark-generated Sn layer retains structural integrity and stable hydrophilicity under repeated liquid exposure, supporting its suitability for long-term photovoltaic applications.

Fig. 10
Fig. 10The alternative text for this image may have been generated using AI.
Full size image

SEM images of the Sn nanoparticle coating (a) before and (b–d) after water-impact durability testing with 1,000, 5,000, and 10,000 droplet impacts, respectively.

Conclusion

This study demonstrates that magnetic-field-assisted spark ablation provides a rapid, scalable, and solvent-free method for fabricating multifunctional SnO2–SnO nanoparticle coatings for photovoltaic applications. Aligning the magnetic field parallel to the substrate is associated with enhanced nanoparticle accumulation and modified phase distribution, leading to the formation of a mixed SnO2–SnO structure under magnetic assistance. Structural analyses (XRD, Raman, TEM/HRTEM) confirm the coexistence of SnO2 and SnO domains, forming a biphasic architecture that may facilitate interfacial charge separation under illumination.

The magnetically assisted coatings show measurable improvements across multiple functional metrics. The ultrathin nanoparticle layer reduces optical reflectance and slightly increases visible-light transmittance, resulting in a 4.02% increase in power output under clean conditions. The superhydrophilic surface enables rapid water spreading and more effective soil removal, reducing performance loss after muddy-water exposure (2.57% compared with 4.21% for uncoated panels), corresponding to a net performance advantage of 6%. The coating also maintains structural integrity after 10,000 water-impact cycles.

By combining material substitution from Ti to Sn, which reduces the required coating passes from three to one, with magnetic-field-assisted high-speed scanning (40 to 70 mm min-1), the overall coating throughput is improved by approximately 5.25-fold relative to baseline spark processing. These results support magnetic-field-controlled spark ablation as a practical approach for large-area, high-throughput fabrication of antireflective, self-cleaning, and anti-soiling photovoltaic surfaces.