Scaling of ultrashort-pulsed laser structuring processes for electromobility applications using a spatial light modulator

Lithium-ion batteries (LIBs) and hydrogen-powered polymer electrolyte membrane fuel cells (PEMFCs) are considered key technologies for transforming the energy and the mobility sector. Both technologies still have restricted performance characteristics, manifested in a limited fast-charging capability for LIBs and a limited reachable current density for PEMFCs. Laser structuring of electrodes for LIBs and the diffusion media (DM) for PEMFCs showed promising results within the literature to overcome the limitations.¹ So far, neither for the LIB electrodes nor for the DM of PEMFCs, the structuring process has been transferred from a laboratory to an industrially relevant scale due to the limited structuring rates achievable by the available laser beam sources and scanning optics. However, recent advances in the development of ultrashort-pulsed (USP) laser beam sources with average output powers in the kilowatt range² and pulse repetition rates in the upper megahertz or even gigahertz region provide new opportunities for industrial applications.³ Furthermore, various novel approaches of multibeam parallelization in combination with USP high-power laser beam sources offer the possibility to achieve industrially relevant processing rates. In the literature, it was reported that the use of polygon scanners,⁴ direct laser interference patterning,⁵ and diffractive optical elements⁶ increased the productivity of material structuring tasks. Beam shaping using a spatial light modulator (SLM) is a promising approach to improve the speed of USP laser processing of large material surfaces⁷ while maintaining a high degree of flexibility with respect to the introduced microstructure. The application of an SLM for the laser structuring of LIB electrodes and PEMFC DM has not yet been studied.

In Fig. 1(a), a schematic representation of an LIB is shown. LIBs have become the market-leading technology for electrochemical energy storage in the fields of consumer electronics, electric vehicles, and stationary systems.⁸ In this context, especially for the application in the mobility sector, LIBs must fulfill high requirements, such as a fast-charging capability and long driving ranges.⁹ To meet these demands, both the volumetric and gravimetric energy density, as well as the power delivery of the cell, must be increased.¹⁰ The resulting conflict of objectives largely depends on the electrode design of an LIB. To achieve a higher cell capacity, the design of the electrode can be adapted by reducing the porosity (defined as the fraction of void volume within the electrode material), by decreasing the mass fraction of inactive materials, or by increasing the coating thickness.¹¹ However, electrodes with a high active material loading and a low porosity show a raised internal cell resistance due to the limited diffusion speed of the lithium ions and, consequently, a loss of performance, especially at higher charging and discharging currents.¹⁰ A promising approach to improve the diffusion kinetics of electrodes is the introduction of three-dimensional structures using laser radiation to reduce the length of the diffusion paths for the lithium ions.¹² In the literature, electrode structuring has been investigated on both the anode (e.g., graphite) and the cathode (e.g., lithium nickel manganese cobalt oxide, NMC) side. However, due to the arrangement of the platelet-shaped graphite particles, a graphite anode has a higher tortuosity perpendicular to the current collector than, for example, an NMC-based cathode.^13,14 Due to the increased diffusion limitation by the graphite particles, promising results were achieved by introducing microstructures into the anode coating. Habedank et al. investigated the effect on the cell performance by drilling microholes with laser radiation into a graphite anode with a hole diameter of about 25 μm and a depth of more than 50% of the material thickness.¹⁰ The microstructures allowed an improvement in the discharge capacity of up to 20% at current rates above 1 C. Furthermore, Hille et al. demonstrated a superior rate capability in discharge tests, particularly at currents between 1 and 3 C, by laser structuring a graphite anode with microholes.¹⁵ In addition to anode structuring, performance enhancements through the microstructuring of various cathode materials have also been reported in the literature.^16–19 

A schematic illustration of a PEMFC is shown in Fig. 1(b). In a PEMFC, the continuous supply of the reactants and an advanced water management, which ensures a consistently humidified polymer membrane and the simultaneous removal of excess water, are essential for an efficient operation.²⁰ The DM, located between the catalyst-coated membrane and the bipolar/end plate, must ensure that the previously mentioned requirements are met. Usually, the DM is a two-layered material consisting of a microporous carbon black-particle-based layer (MPL) and a macroporous carbon fiber-based substrate layer, commonly referred to as a gas diffusion layer (GDL).²⁰ In various studies, it has been shown that liquid water accumulates in the pores of the DM at high current densities (typically 1–2 A/cm²), preventing the oxygen from reaching the catalyst layer, which resulted in a higher overpotential and a reduced output voltage.²¹ Microholes in the DM can improve the water removal significantly and increase the power density of the fuel cell.²² So far, it has been shown that introducing microholes at different positions and with different geometries can have beneficial effects on the fuel cell performance. Gerteisen et al. demonstrated that microholes with a diameter of 80 μm in the GDL led to an increase in the limiting current density of 20%,²² while Wang et al. observed the best fuel cell performance with microholes of 100 μm.²³ The perforation of both layers (MPL + GDL) showed advantages in various studies. However, no definite microhole size has yet emerged. Alink et al. demonstrated that holes with a diameter of 80 μm led to improvements in the fuel cell performance,²⁴ while Haußmann et al. figured out that a diameter of 60 μm is advantageous.²⁵ Additionally, it has been demonstrated that introducing microchannels into the MPL can have beneficial effects on the water transport under certain operating conditions.²⁶ 

Both components, LIB electrodes and PEMFC DM, are produced in continuous roll-to-roll coating processes. Within the electrode production, the active material slurry is coated with a doctor blade or slot-die application tool on the current collector foil.⁷ In an industrial process, the material passes through a drying section to remove the solvent²⁷ with high velocities of approximately 1 m/s.⁶ With a coating width of 1 m and a hole distance of 200 μm, this results in a structuring speed of 25 × 10⁶ microholes per second. Standard coating thicknesses are in the range between 50 and 100 μm.²⁸ 

For the production of the DM, the MPL slurry is coated onto the GDL with a subsequent sintering process.²⁹ Due to a significantly lower demand of fuel cells at the moment, it is assumed that the production speed is one order of magnitude smaller than that of the LIB electrodes at approximately 0.1 m/s and the hole distance is between 500 μm and 1 mm. These values result at a coating width of 1 m to a structuring speed of 200 000 or 400 000 holes per second. For commercially available DM, the MPL thickness is in the area of 28–134 μm.³⁰ 

Due to the design of the industrial production process for the electrodes and the DM, the microstructuring of the materials must be integrated into the roll-to-roll process and meet state-of-the-art production speeds to ensure the economic efficiency of this additional material processing step.

In order to increase the speed of the laser structuring processes, various concepts have been investigated:³¹ 

• Increasing the scanning speeds with ultrafast scanning systems, such as polygon scanners, at optimized laser beam fluences and the highest possible pulse repetition rates (PRR); and

• Increasing the processing speed by dividing the laser beam into multiple parallel beams with optimized fluences for each sub-beam and with lower PRRs to minimize heat accumulation (kilohertz region).

So far, it has been shown that for structuring LIB electrodes or the DM of PEMFC, more than one pulse is needed to obtain a sufficient depth of the microhole.^10,32 Using a polygon scanner to perform the structuring process within a continuous roll-to-toll process would pose the challenge to hit a microhole multiple times with separate microholes. This would demand a precise synchronization between the scanner and the roll-to-roll machine.^4,33

Yamada et al. developed a high-precision and efficient laser system specifically for the mass production of through-holes in anodes and cathodes of lithium-ion batteries. The setup consists of a hollow cylinder around which the electrodes are wound, whereby a pulsed laser beam, which is guided through the central axis of the cylinder and reflected radially through precise openings, continuously creates evenly distributed through-holes in the electrode layers.³⁴ 

For the beam splitting, various options exist, such as DOEs⁶ or microlens arrays.³⁵ These technologies provide a fixed intensity distribution and number of sub-beams.³⁶ Hille et al. used a diffractive optical element to split the laser beam into 21 sub-beams with a 3 × 7 configuration. During the laser structuring of LIB electrodes, a decrease in the processing time by 88% was observed in comparison to the electrode structuring using a single-beam process. The microholes, created either during the multibeam or single-beam process, showed similar geometries and mechanical integrities.⁶ Within the study of Hauschwitz et al., the laser beam was divided into 40 401 sub-beams to create a superhydrophobic surface on a stainless-steel plate. With a moving stage, a processing rate of 8 × 10⁶ microstructures per second is expected to be achievable.³⁷ 

In addition to optical elements that generate fixed output power distributions, some systems can change the output power distribution and number of sub-beams at a certain frequency.^36,38 These systems are SLMs. SLMs are adaptive optical devices capable of modulating the characteristics of an incident optical wavefront, including the phase, the amplitude, or the polarization.³⁹ 

A prevalent form of an SLM is the liquid crystal-(LC-)based SLM. Such a device operates on the principle that the phase change of light polarized along the extraordinary axis of the crystal can be precisely regulated by the variable refractive index of LC materials. LC-based SLMs typically modulate the light in response to either optical or electrical inputs. An optically addressed SLM features a continuous photosensitive layer atop the modulating material, enabling the modulation of a light beam using another beam. However, its manufacturing costs are relatively high. Conversely, an electrically addressed SLM consists of a pixelated structure manipulated electrically to modulate the local optical wavefront at each pixel. While this type of device offers a direct interface between the optical and the electronic units, drawbacks include lower light utilization efficiency due to dead zones between the electrodes and diffraction losses from the pixelated structure.³⁹ 

Silvennoinen et al. showed the structuring of stainless steel and silicon with a femtosecond-pulsed laser beam separated into 576 sub-beams by an SLM.⁴⁰ Within a subsequent study from Silvennoinen et al., it was demonstrated that an SLM could be used to divide a laser beam into 400 sub-beams with varying intensity distributions to structure a silicon wafer.⁴¹ Lutz et al. investigated the effects of the parallelized laser beam processing using an SLM on the ablation behavior and the resulting surface roughness of stainless steel as a function of the fluence. It was demonstrated that beam splitting up to 20 sub-beams can be used to increase the ablation rate. At 20 sub-beams, a smaller ablation efficiency was observed, which was explained by a decreased homogeneity between the sub-beams and a decreased diffraction efficiency by the SLM.⁴² In the study by Yoshizaki et al., it was demonstrated that an SLM can be employed to divide a single laser beam into three sub-beams of varying diameters, thus enhancing the processing speed in glass structuring applications.⁴³ 

This work investigated the feasibility of scaling the laser structuring of LIB electrodes and DM for PEMFCs by flexible laser beam shaping. An SLM was applied for the parallel introduction of microholes and channels into the materials. To evaluate the applicability of an SLM for laser microstructuring, the ablation quality obtained by employing the SLM was compared with the results of a single laser beam deflected by a scanning optics.

The experimental investigations were conducted using commercially available DM (SIGRACET 36BB, SGL FCC GmbH, Germany) and LIB graphite anodes and cathodes (UniverCell Holding GmbH, Germany). The characteristics of both material systems are summarized in Table I.

For the structuring experiments, a USP laser beam source was used (Perla, HiLASE Centre, Czech). The system emits laser pulses with a duration of 970 fs at a central emission wavelength of 1030 nm. For beam splitting and shaping, a liquid crystal on silicon SLM (X15223, Hamamatsu Photonics K.K., Japan) was integrated into a flexible beam-shaping unit (FBS G3, Pulsar Photonics GmbH, Germany). The beam deflection was performed by a 2D scanning optics (IntelliScan 14, Scanlab GmbH, Germany). The system used was described in more detail by Hauschwitz et al.⁴⁴ In Table II, a summary of the laser and the optical setup characteristics is shown.

To investigate the impact of the single- and the multibeam approaches for the anode structuring on the performance of the battery cell, full coin cells were assembled and characterized in discharge rate capability tests. The discharge rate capability tests were performed similar to tests shown already by Habedank et al. and Hille et al.^10,15 The entire cell assembly process was conducted in a dry room with a dew point of approximately −40 °C. First, the electrodes were dried in a vacuum oven (Goldbrunn 1450, Goldbrunn Therm GmbH, Germany) at 120 °C and 50 mbar for 24 h to extract the residual moisture. The 2032-type coin cells consisted of an anode (15 mm diameter) and a cathode (14 mm diameter), which were separated by a glass fiber separator (16 mm diameter, type 691, VWR, USA). The employed cathode material is listed in the supplementary material. For the coin cell filling, 100 μl electrolyte (LP572, BASF SE, Germany) was used, containing ethylene carbonate and ethyl methyl carbonate (mass ratio of 3:7) with 1 mol lithium hexafluorophosphate as the conducting salt and 2 wt. % vinylene carbonate. Four cells were built for each laser parameter configuration and setup listed in the Appendix A of the supplementary material, as well as for the reference cell without the structured anode. After the cell assembly, the cell formation was initially conducted with a battery cell test system (CTS, BaSyTec, Germany) at a constant temperature of 25 °C. For the formation, three charge and discharge cycles were performed with defined C-rates. The C-rate describes the ratio of the charging or discharging current in amperes (A) to the capacity of the cell in ampere hours (Ah). The formation was conducted at a constant current of 0.2 C between 2.9 and 4.2 V. The third charging cycle was followed by a constant voltage phase at 4.2 V until the current dropped below 0.05 C. Subsequently, for each cell configuration, rate capability tests were performed in which the cells were charged and discharged between 2.9 and 4.2 V at discharge rates of 0.1 to 5 C. Only discharge rate tests were performed to avoid lithium plating and, thus, the risk of test falsification for subsequent test cycles. A more detailed test description and the entire test protocol can be found in the Appendix E of the supplementary material.

For a better comparison of the structure geometries, the influence of the additional losses caused by the optical path and the SLM was measured before the experiments for this study were conducted. A power loss of 7% due to the optical setup and an additional 15% could be determined when an SLM was used for beam shaping/splitting. Thus, during the experiments, the reported pulse energies and output powers were those measured at the workpiece, accounting for the power losses.

In the remainder of this publication, the term “structuring” will refer to the percussion drilling process used to create blind holes. All materials were structured on one side. For all structuring experiments, a PRR of 50 kHz was chosen. In the first study, the pulse energy was set to values of 80 and 100 μJ while simultaneously changing the number of pules for each hole to keep the total energy dose per microhole constant at 2 and 4 mJ for the DM and the anode, respectively. Considerably high pulse energies were used, as previous studies have shown that a higher ablation depth can be reached per pulse in this way.^6,32 Reducing the number of pulses per hole is assumed to facilitate the structuring task in a roll-to-roll production process, where the introduction of microholes into the continuously moving DM or LIB electrode material requires precisely tuned control systems. At first, the microholes were created sequentially using a single beam moved by the scanning optics. In the second step, the SLM was applied to split the beam into a 2 × 2 matrix, keeping the pulse energy and the hole energy constant compared to the previous experiments. The two different strategies to introduce the microholes are schematically shown in Fig. 1(a) and Fig. 2(b), where d_f equals the focal diameter of the laser beam in the focus spot of ≈30 μm. The spot size of the sub-beams generated by the SLM was comparable to the spot size of the original single beam. This occurred because the SLM divided the raw beam into sub-beams with dimensions similar to those of the raw beam. The distance x between the holes was set to 1 mm²⁵ for the DM and 200 μm for the anodes,⁶ which was based on values that showed promising performance improvement in the literature.

The intensity distribution of the Gaussian laser beam measured after the SLM within the optical path can be seen in Fig. 3(a). In Fig. 3(b), the intensity distribution (measured in the focus point) of the line shape generated by the SLM is shown.

Figure 4 presents the hole geometries of the two materials (DM and anode) using the single-beam as well as the multibeam process. For both the DM and the anode materials, the geometrical characterization of the holes using the LSM indicated no substantial geometric deviations between the single-beam and the multibeam approaches. At a pulse energy of 100 μJ and 40 pulses per hole, the hole structures of the anode using the SLM showed a slightly smaller average diameter compared to the single-beam method, while the value of the average depth was nearly identical for both processes. The reduction in the pulse energy to 80 μJ and the increase in the number of pulses to 50 resulted in a decreased average hole diameter and depth for both methods. However, the geometric deviations between the two structuring processes remained marginal. The DM material displayed a nearly identical average hole diameter and depth at a pulse energy of 100 μJ and 20 pulses for both methods. At 80 μJ and 25 pulses, the use of the SLM resulted in a minimal higher average diameter and depth than the single-beam method. Since the deviations of the average hole diameter and depth between the single-beam and the multibeam approaches for both materials were below the respective standard deviation, it can be concluded that the application of the SLM for the multibeam method did not significantly affect the ablation quality (compare Fig. 5).

Figures 5(a)–5(d) illustrate the ablation characteristics of the graphite particles in the anode for both structuring processes qualitatively after using SEM imaging. For the single-beam as well as for the multibeam approach, the morphology of the holes revealed whole graphite particles that were not perforated by laser radiation. This can be attributed to the ablation behavior of the anode coating, which is initially caused by the evaporation of the binder, creating a force that ejects the particles from the material.^46,47 In addition, graphite does not melt under ambient pressure, causing a fraction of the whole particles to be vaporized directly by the energy input of the laser.^46,48 A similar ablation behavior is assumed for the DM material, but the particle sizes of the applied carbon black are significantly smaller and the polymer binder content is higher compared to conventional battery anodes [compare Figs. 5(e)–5(h)].³² 

To further investigate the geometry of the microholes, the profile through the deepest point was extracted from the LSM measurements, and an average value was calculated for eight microholes (compare Fig. 6). The depiction of the profiles illustrated a similar geometric shape of the holes when comparing the single-beam and the multibeam patterning of the DM and the anode materials.

Figure 7 represents the depth and width values for the microchannels. The ablation depths for the DM are comparable between the two approaches, while the ablation widths of the channels generated by the SLM are smaller. This behavior was observed for both cases. In contrast to the findings for microholes in the DM and the microholes, the anode material showed a deviating behavior when the SLM was employed to insert the microchannels. Compared to the single-beam method, the channels had a significantly smaller depth but a comparable width.

For further evaluation, contour height maps and mean cross sections of exemplary microchannels are presented in Figs. 8(a)–8(i), respectively. This methodology was presented by Kriegler et al. to evaluate and compare the channel quality for microchannels introduced in ceramic materials.⁴⁹ 

For the DM, the structures introduced by the SLM revealed a much more pronounced channel shape in terms of depth and width than the channel geometry of the single-beam method. In addition to the microchannels, also the cracks are visible. In contrast, the channels in the anode material showed the opposite effect, where the single-beam-generated lines had a more defined contour. The cross section plots of the microchannels can be seen in Fig. 8(i). As already apparent in Fig. 7, the channels in the DM created by the SLM approach had a comparable depth but a smaller width. In addition, the cross section exhibited the contrary behavior of the anode material, which was characterized by a significant deterioration of the channel contour due to the SLM structuring.

The significant deviation of the channel geometry between the SLM line shape and the single-beam approach in case of the anode presumably occurred due to the ablation behavior of the anode material and to the intensity distribution of the beam shaping [compare Fig. 3(c)]. When using the line shaping method, the introduced energy was distributed over a larger area compared to the beam shaping used to create the microholes, which led to a lower overall fluence. In addition, the pulse energy was not uniformly distributed over the area of the line, resulting in an increased intensity in the center of the line, which decreased steadily toward the line edges. With regard to the previously explained ablation behavior of the graphite anode, this implied that the application of the line shaping by the SLM did not provide sufficient energy to completely ablate the graphite particles, particularly in deeper layers of the electrode and at the edges of the beam shape where the intensity is lower.

In Fig. 9, the effective permeability determined for an unstructured DM as well as for DMs structured with microholes by a single-beam process and by an SLM multibeam process is shown. The DM with the microholes had a higher effective permeability (1.37 × 10⁻³ and 1.24 × 10⁻³ μm²) compared to the unstructured DM (0.97 × 10⁻⁴ μm²).

After analyzing both DM with the microhole pattern, no significant difference was observed. In addition, the DM with microchannels (single-beam and SLM-generated line) was tested. Both showed a significantly higher permeability compared to the DM with microholes. This could be explained by the improved water transport properties due to the larger structure sizes. Again, no difference was observed between the single-beam process and the SLM line process.

In Fig. 10, the discharge capacities at C-rates ranging from 0.1 to 5 C of the full cells with all structuring approaches tested are depicted. All cells showed a decrease in the discharge capacity by increasing the discharge rates. However, the cells with a structured anode achieved higher capacities at C-rates greater than 1 C than the reference cells. At a C-rate greater than 2 C, the cells with single-beam holes and a pulse energy of 100 μJ retained fractionally higher discharge capacities than the cells containing holes obtained by the SLM. However, since the deviations of the average capacity between the single-beam and multibeam methods were below the respective standard deviation for all C-rates, it could be concluded that the use of the SLM had no significant influence on the performance of the cells.

Particularly, at a pulse energy of 80 μJ, the differences in the capacities between the two structuring methods were negligible throughout all C-rates. In contrast, at a C-rate above 2 C, a clear deviation of the discharge capacities between the single-beam and the SLM method was apparent for the line geometry. This effect could be explained by the previously shown difference in the material ablation between the two methods, as the use of SLM led to a significantly lower ablation depth of the microchannels (compare Fig. 8). The control cycle of 0.1 C at the end of the discharge rate test showed a slight loss of capacity for all cells, although the loss was most pronounced for the reference cells. However, the differences between the structuring methods were marginal.

To enable an economical industrial application of laser structuring, the process must be integrated into the roll-to-roll manufacturing processes of battery electrode and DM production. The substantial material throughput related to these production steps mandates the structuring process to be executed at elevated velocities and rates. Hille et al. calculated the output power needed to structure battery anodes at industrial processing rates and with suited microhole distances to be in the kilowatt regime.⁶ In contrast to the battery electrodes, no studies have been carried out yet on scaling for DM, which is why the scope of this section is on this material. The schematic structure of the model used to examine the scaling of the structuring process is presented in the supplementary material. In the calculation, three different line speeds of the roll-to-roll machine were assumed (1, 15, and 30 m/min). As the structured area, the working field of a conventional galvanometer scanning optics system used for micromaterial processing was selected (75 × 75 mm). In this case, this means that if a wider material needs to be structured, the number of scanning optics working in parallel must be increased. For the calculation, the scanner parameters, such as the delay times and moving speed, were considered as well (compare the supplementary material). For the process parameters, the pulse energy 80 μJ and 25 pulses per hole at a PRR of 50 kHz at a maximum were chosen. Figure 11 shows the average output power needed for laser structuring of microholes introduced with different spaces between the holes and with different beam arrangements. At a line speed of 1 m/min, the single-beam process has its limit at a hole distance of 1.5 mm. For the smaller hole distances, the process is too slow, and the structuring of the entire area is not possible once it has been moved out of the working area of the scanning optics.

For a line speed of 15 m/min and a hole distance of 0.2 mm with a 40 × 40 matrix, a higher average output power is necessary than structuring with a matrix consisting of more sub-beams. This observation can be explained with the fact that fewer scanner movements with the associated on-and-off delay times are necessary when larger beam matrices are used, as a wider DM area is covered. The time saved is additionally available for drilling the holes, resulting in a smaller average output power. As the structured area was limited to the working area of one scanner system, it can be assumed that structuring wider areas leads to a linear increase in the needed output power. If, for example, a DM with a width of 0.5 m should be structured with a 0.5 mm spacing between the holes at a line speed of 15 m/min, a laser source with an average output power of 250 W is necessary. In the scenario under consideration, the required output power is within the capabilities of commercially available USP beam sources. However, augmenting the line speed, reducing the spacing between microholes, or increasing the energy per hole to enhance the structure depth will lead to a substantial increase in the required output power.

In addition to the required output power of the laser beam source, the damage threshold of the SLM must be taken into account. For the SLM used, the damage threshold for femtosecond laser pulses is at a pulse energy of 1 mJ and at an average output power of 100 W. To overcome these limitations, the high-energy pulses can be split into bursts. Alternatively, increasing the PRR of the laser while reducing the pulse energy is another viable approach. Finally, it can be assumed that the damage thresholds for SLMs will gradually increase for femtosecond laser pulses, similar to the trend observed with continuous wave laser radiation, where SLMs are currently capable of handling output powers up to 700 W.⁵⁰ 

The results obtained in this study confirmed that applying an SLM enables high-quality microstructures with significantly higher structuring speeds compared to a conventional single-beam approach. Hence, the implementation of this beam-shaping technology holds the potential to contribute to the transfer of the laser structuring from a laboratory to an industrial application for the battery and fuel cell production. Besides the increase in the structuring rates due to beam splitting, the potential of the high flexibility in changing the beam shapes or the size and the amount of the sub-beams within milliseconds is promising. In the DM production, this flexibility allows to produce products with varying properties within one single production line and one production load.

Additionally, a clear difference in the ablation behavior of LIB anodes and PEMFC DM, especially for the microchannels, was visible. This behavior could be attributed to the different materials as well as material compositions used. Additionally, the inhomogeneous intensity distribution of the shaped beam was presumably a reason for the observed differences. This could change the ablation characteristics especially at the outer parts of the beam profile.

With regard to the scaling of the process, it becomes evident that beam splitting is necessary to increase the velocities needed in the roll-to-roll processes.

Further studies should investigate how the possibility of changing and homogenizing the intensity distribution of the single-beams and the created beam shapes by the SLM can improve the geometry of the created structures. In particular, the line geometry generated by the SLM requires further optimization for different material systems, as significant deviations between the anode and the DM were evident. Additionally, the losses occurring through the use of the SLM have to be further investigated and reduced in order to pave the way for a further commercialization.

See supplementary material for additional information on various aspects of this study. This includes a description of the experimental design employed in the laser structuring studies (refer to Appendix A) and the construction and utilization of the test bench for determining the effective permeability (refer to Appendix B). Additionally, the supplementary material covers the boundary conditions, parameters, and the schematic structure of the model used for the process scaling calculations (refer to Appendix C). Detailed information regarding the cathode material used in the coin cells (refer to Appendix D) and the protocol for the formation and rate tests of the lithium-ion batteries (refer to Appendix E) are also provided.

The results presented were achieved within the research project Impedia (No. 03B11045A), which is supported by the German Federal Ministry for Digital and Transport (BMDV) within the funding program National Innovation Program Hydrogen and Fuel Cell Technology Phase II. The program is coordinated by the NOW GmbH and the project was supervised by the Project Management Organization Jülich (PTJ). This work was co-funded by the European Union and the state budget of the Czech Republic under the project LasApp CZ.02.01.01/00/22_008/0004573. The authors would like to thank the BMDV, the NOW GmbH, the PTJ, the European Union, and the Czech Republic for their support and trusting cooperation.

The authors have no conflicts to disclose.

Christian Geiger: Conceptualization (equal); Investigation (equal); Methodology (equal); Project administration (equal); Visualization (lead); Writing – original draft (equal); Writing – review & editing (equal). Alena Gruendl: Conceptualization (equal); Investigation (equal); Methodology (equal); Project administration (equal); Visualization (equal); Writing – original draft (equal); Writing – review & editing (lead). Petr Hauschwitz: Investigation (equal); Resources (equal); Writing – review & editing (supporting). Ivan Tarant: Investigation (equal); Resources (equal); Writing – review & editing (supporting). Lucas Hille: Conceptualization (equal); Investigation (supporting); Methodology (equal); Writing – review & editing (supporting). Alessandro Sommer: Investigation (equal); Visualization (supporting); Writing – review & editing (supporting). Bolin Hou: Investigation (equal); Visualization (supporting); Writing – review & editing (supporting). Michael F. Zaeh: Funding acquisition (equal); Project administration (equal); Supervision (lead); Writing – review & editing (equal).