Open Access
Add to BibSonomy
Abstract
The bursts of picosecond laser pulses have nanosecond-level short interval delay. These bursts contain a variable number of sub-pulses, which are used for laser cutting of copper current collector and graphite anode material for Li-ion battery anode. The influences of 2–10 sub-pulses on kerf edges were studied and were compared with that of a single pulse. The shapes of anode edge cut under different conditions, obtained using scanning electron microscopy (SEM), revealed that using burst mode would yield a smaller heat-affected zone (HAZ) of the copper current collector and smaller delamination width of graphite anode material. The capability of laser cutting of anode was characterized with maximum single-time cutting speed. Results showed that the cutting efficiency was raised evidently with the increase in the number of pulses in a burst, and the maximum cutting speeds for the copper current collector and graphite anode material could reach 3,800 mm/s and 500 mm/s respectively.
© 2021 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Characterized by high energy density [1], high voltage, low self-discharge rate [2], no memory effect [3], and long cycle life [4], etc., Li-ion batteries are ideal power sources for portable electronic devices and emerging unmanned aerial vehicle (UAV) products and also preferred power sources for electric and hybrid vehicles. A Li-ion battery is composed of a copper anode coated with graphite active material, an aluminum cathode coated with lithium iron phosphate (LiFePO4), separator film and electrolyte, and its production is completed through cutting, stacking, electrolyte filling, and packaging [5,6]. The current large-scale and wide applications of Li-ion batteries require more flexible production means to adapt to varied sizes of Li-ion battery products, and laser cutting provides a very good solution. In order to overcome such problems in mechanical blanking as restriction to product compatibility by die, long time for model changeover, and stress-caused stretching of, curling of and active material coming-off from electrode [7], the cutting process for Li-ion battery begins to turn to laser cutting as a new alternative. The laser cutting of electrode is non-contact machining, and it has the advantages that the demand for the shape of a battery product can be met through model changeover by one key and the product quality can be modified and optimized by adjusting the laser parameters. Moreover, a laser cutting system can operate continuously in the production line, with small vibration and low subsequent maintenance cost.
Cutting of Li-ion battery electrodes with laser irradiation has recently gradually become a research hot spot, which generally focuses on two aspects, i.e., cutting quality and cutting efficiency. The comparison of research efforts relating to the maximum cutting speed and cutting quality for electrode material cutting with continuous laser and long pulse laser [8] shows that high power continuous laser can realize high speed cutting, whereas pulsed laser is superior in the research efforts focusing on edge quality [6]. After laser cutting of electrode material, there are mainly problems of HAZ and delamination. High heat machining mode will cause oxidation of base material. The aluminum oxide formed on the cathode does not conduct electricity, so it will not influence the performance of battery. However, the copper oxide formed on the anode will act as a discharge cathode, so it will influence the performance of battery. Lee et al. developed a 3D mathematical model for high-speed laser cutting of current collector, and had proven that the cutting of aluminum current collector relied on laser intensity and the cutting of copper current collector relied on laser intensity and interaction process [9,10]. In a sandwich structure with active material coated on current collector, the delamination phenomenon that the metal current collector is bared without coating after cutting reduces the effective area of the electrode, resulting in decrease in battery capacity. It has been found that the kerf depth in an electrode with sandwich structure is a piecewise function of mean laser power and the metal conductor layer plays a leading role in the cutting process [11,12]. Schmieder [13] proposed a model of theoretical ablation mechanism for electrode surface material. Lutey et al. studied the high speed laser cutting of electrode for Li-ion battery, providing guidance for selection of process parameters to realize optimal process efficiency and visible cutting quality with a high power ns-pulsed laser source, and proposed that shorter laser pulses could further reduce cutting energy input and improve cutting quality [14]. Until now, most research efforts on laser cutting of electrode for Li-ion battery have focused on continuous laser and long pulse laser, and there have been few ones on ultrafast laser cutting of electrode for Li-ion battery. Pfleging [15] et al. demonstrated through chemical analysis with time-of-flight secondary ion mass spectrometry (TOF-SIMS) that a thin copper contamination layer, that is, the delamination phenomenon of copper metal conductor layer baring, was detected on the top of anode electrode undergoing nanosecond laser cutting, but the delamination could be completely eliminated by ultrafast laser cutting with pulse length of 350 fs. It was also reported that the delamination could also be completely eliminated by multiple times of scanning cutting with femtosecond ultrafast laser [12]. Of course, it is impractical to use femtosecond laser technology in industrial production due to the cost and efficiency.
According to the current research situations, there are following problems in laser cutting of electrode for Li-ion battery: 1) it is hard to obtain high quality kerf edges through cutting with long pulse laser; 2) if femtosecond laser cutting is used to obtain high quality kerf edges, the invested cost is too high, so it is difficult to use femtosecond laser cutting in practical industrial production; and 3) the cutting efficiency in the current research efforts relies on high power ultrafast pulsed laser. These make laser cutting developed slowly in the cutting of electrode for Li-ion battery. For this reason, burst mode was introduced for cutting based on the process of laser cutting of electrode for Li-ion battery. In the burst mode, bursts with very high repetition frequency can be generated every time when the laser is emitted. When the interval time between the sub-pulses in a burst is shorter than the thermal relaxation time of the target material, the ablation cooling [16] can be realized, and there is no evident thermal diffusion occurring before the next sub-pulse arrives, so the heating is reduced in the surrounding zones. Compared with conventional single pulse laser irradiation, the emerging burst mode enlarges the adjustable room for laser machining technology, more favorable for better optimized process effect. At present, the use and study of burst mode have been carried out in micromachining of metals [17–19]. There have been researched results showing that the burst mode can raise the ablation efficiency [20], increase the material removal volume [21], reduce the surface roughness [22], and so on. However, there are currently few reports on the cutting of electrode for Li-ion battery in burst mode.
In this work, we studied the edge cutting quality and cutting efficiency by cutting copper current collector and graphite anode material for Li-ion battery anode with a variable number of picosecond laser pulses in burst mode. We obtained the kerf edges of electrode under different conditions and the maximum cutting speeds with different parameters by changing the laser power and the number of pulses in a burst. We measured the microscopic morphological characteristics of the kerf edges of electrode using the SEM, and discussed the scope of HAZ through the element composition variations in the kerf edges obtained through energy dispersive X-ray spectroscopy (EDX). The purpose of this paper is to provide an electrode cutting solution with low power ultrafast laser irradiation, which can meet the industrial production requirements.
2. Materials and methods
2.1 Electrode sample
A commercial Li-ion anode was used in this experiment, and it included a 6 µm extremely-thin copper metal current collector and active graphite anode material coated on both sides with equal thickness. The thickness of each graphite coating layer was 60 µm, and the total thickness of the sandwich structure of the anode was 126 µm. The anode is shown in Fig. 1.
2.2 Experimental setup
A Hymson BR-PS-50-1064-W infrared picosecond solid-state laser with emission wavelength of 1,064 nm, maximum output power of 50 W and pulse duration ≤10 ps was used in the experiment. The laser could operate in single pulse mode or burst mode. Pulses were generated at an interval of t in the single pulse mode, and densely-distributed pulse envelopes were generated in the burst mode. In one pulse envelope, the interval between pulses was expressed with tp, as shown in Fig. 2. In the experiment, t=2.5 µs, and tp=30 ns. The number of pulses in a burst was set from 2 to 10 in the experiment, which represented the number (N) of generated corresponding sub-pulses. The spatial intensity distribution of sub-pulses was approximately an ideal Gauss distribution. In addition, the peak energy of burst decreased with the increase in the number of pulses in a burst. The beam quality factor of the laser was M2 ≤1.3, and the laser irradiation was focused onto the material surface with a telecentric lens with the focal length of 167 mm through a 4x beam expander, to obtain spot diameter of 42 µm.
2.3 Experimental method and process
The electrode material was placed on a workbench. There was a kerf under the cutting area of the workbench to avoid that the sample came into contact with the workbench to have an unnecessary reaction during cutting. The copper current collector and graphite anode material were allowed to undergo single scanning exposure at the focus, and it was ensured in the scanning every time that the electrode was completely penetrated without adhesion in the kerf. In the case of a single pulse, the preliminary test was conducted with the laser power decreased from 95% at an interval of 10%, to determine the minimum cutting powers for the copper current collector and graphite anode material. It was found that the minimum cutting powers for copper current collector and graphite anode material were 34.2 W and 31.2 W respectively. Afterwards, the electrode materials underwent exposure to laser with the number of pulses in a burst ranging from 2 to 10 at a fixed galvanometric scanning speed of 600 mm/s, and the minimum cutting speed in single scanning mode was studied under different laser powers and different numbers of pulses in a burst.
The kerf edges were observed and analyzed with a Keyence VHX-6000 optical microscope, and the micromorphology of kerf edge was further characterized through the SEM. The cutting quality was compared between single pulse scanning mode and burst mode. The surface stoichiometry of the copper current collector after exposure to laser was analyzed through EDX, to determine the scope of HAZ.
3. Results and discussion
3.1 Laser cutting of copper current collector
Figure 3 shows the HAZ width in the kerf edge of copper current collector with varied numbers of pulses in a burst, and the width change trend of white border occurring in the scope of HAZ. In the single pulse mode, the heat affected phenomenon in the kerf edge was reflected as different discoloration regions. The near kerf end obtained very high heat, exhibiting a white border region with bright white color, followed by blue discoloration region and red thermal diffusion region. The color faded gradually to the original color of copper current collector with the increase in distance to the kerf edge. In the burst mode, when N = 2, the HAZ increased instead, which could be attributed to the plasma shielding effect [12,16,23]. Two split-formed sub-pulses had relatively high incident fluences, and the cloud generated by the radiation of the first pulse to the copper absorbed the energy of the second pulse that followed, resulting in the re-deposition and re-diffusion of heat in the copper. When N = 3, it could be considered that the shielding effect had been eliminated. Afterwards, when the number of pulses in a burst was increased gradually, the white border width decreased significantly. When the number of pulses in a burst was increased to N = 4 and above, the white border was completely eliminated, and the whole HAZ in the kerf edge started to exhibit a decreasing trend until the number of pulses in a burst reached N = 6. At a fixed scanning speed of 600 mm/s, the copper current collector had the smallest HAZ width, ∼35 µm, when the number of pulses in a burst was N = 8.
Fig. 3. Relationship of HAZ in copper current collector with number (N) of pulses in a burst at 600 mm/s, 400 kHz and 34.2 W (the case with number of pulses in a burst of 1 was regarded as single pulse mode). The illustrations show the optical microscopic images of kerf edge when the single pulse mode N=1 and the burst mode N=8.
As shown in Fig. 4, the micromorphology of the kerf in copper current collector at 600 mm/s and 34.2 W was characterized through the SEM. After exposure in single pulse mode, no recast phenomenon was observed at the kerf edge, the edge was not smooth, and there were irregular burrs with size of fewer than 10 µm distributed at the kerf. In addition, there was a large amount of debris deposited on the surface of copper current collector, and even more, ones deposited at the burrs and kerf. It can be seen that the cutting process in the single pulse mode was accompanied by the occurrence of phase explosion, which caused a large amount of liquid and gas mixtures to splash out. In the burst mode, the smallest HAZ was obtained when N = 8. As can be seen from the SEM image, the kerf edge was relatively clean, only with a little splashing material adhering. There was a recast layer with a width of ∼10 µm at the kerf, which indicated that the mechanism of cutting of copper current collector was changed in the burst mode. Owing to the short duration of picosecond laser pulse, which reached the electron-phonon relaxation time, and the nanosecond-level short time delay between the sub-pulses in the burst mode, the heating in the surrounding zones was further reduced. During cutting, the copper current collector in the laser pulse irradiation zone was heated to extremely high temperature within an extremely short time, and then was removed directly by rapid thermal gasification and took away most of the heat. A little residual heat melted the copper, which then resolidified to form a recast layer at the kerf. After the number of pulses in a burst was more than 8, the HAZ increased with the increase in the number of pulses in a burst (as shown in Fig. 3), which was caused by the accumulated residual heat in the material.
Fig. 4. SEM images of kerf edge of copper current collector in (a) single pulse mode and (b) burst mode with N = 8.
Figure 5 shows the maximum cutting speed Vm for copper current collector obtained at different numbers of pulses in a burst and different laser powers. The Vm increased with the increase in the number of pulses in a burst, and higher laser power corresponded to higher Vm. Increasing the number of pulses while decreasing the pulse energy was favorable for avoiding the plasma shielding effect, to raise the cutting efficiency. The maximum Vm of 3,800 mm/s was obtained at 37.4 W and the number of pulses in a burst N=10, and the cutting speed was increased by over 5 times, compared with that in the single pulse mode. The optical microscopic image of kerf edge obtained at this parameter value showed that the edge color was almost consistent with the original color of the substrate and there were no HAZ with color change and evident defects. The maximum cutting speed showed that the cutting efficiency for copper current collector was influenced by the number of pulses in a burst and the laser power. However, it can be seen that the increase in maximum cutting speed caused by the increase in the number of pulses in a burst was evidently larger than that caused by the increase in laser power. A picosecond laser with maximum power of 50 W was used in the experiment, and its power increase range was limited. The burst mode could realize rapid machining with a low power laser.
Fig. 5. Maximum cutting speed for copper current collector at different numbers of pulses in a burst and different laser powers, and kerf morphology at number of pulses in a burst of 10 and laser power of 37.4 W.
The micromorphology obtained at maximum Vm of 3,800 mm/s in the experiment is shown in Fig. 6(a). There was a recast region of 15 µm at the kerf. Under high speed cutting condition, the recast region exhibited scale-like stacking and was accompanied by a few splashing particles adhering to the copper surface. In order to determine the scope of the HAZ, the stoichiometry of the kerf edge was further analyzed through EDX to observe the content changes of elements Cu and O in the kerf edge. The line scanning analysis results in Fig. 6(b) showed that the element O content fluctuated in the scope of the recast layer, demonstrating that oxidation reaction occurred between the recast layer and the oxygen in the air during solidification. Outside the recast region, the stoichiometry of various elements did not have an evident change, indicating that the HAZ at the kerf edge was limited within the recast region.
Fig. 6. (a) SEM image of kerf edge of copper current collector at number of pulses in a burst N=10,3800 mm/s and laser power of 37.4 W, where the red arrow denotes the EDX measurement position, (b) EDX line scanning analysis results.
3.2 Laser cutting of graphite anode material
When graphite anode material was cut with laser, delamination phenomenon could be observed at the kerf edge, since the laser absorptivity and thermal diffusion rate are different between the middle copper conductor layer and the graphite coating layers on both sides. Figure 7 shows the kerf morphology of anode at 200 mm/s and 34.2 W characterized through the SEM. After exposure in the single pulse mode, delamination width of 70 µm occurred in the upper graphite layer, with a large amount of debris adhering to the kerf edge and without recast phenomenon occurring. When in the burst mode with N = 6, the delamination width decreased, remelting phenomenon occurred in the middle copper layer at the kerf, and the debris deposition was severe. When the number of pulses in a burst was increased to N = 10, there was an evident recast layer, the deposited debris reduced and the delamination width decreased to 35 µm. As can be discerned, the delamination width decreased with the increase in the number of pulses in a burst. The laser cutting of graphite anode with sandwich structure was mainly influenced by the middle copper conductor layer. Since copper had high light absorptivity near the wavelength of 1 µm, when laser radiation reached the copper conductor layer, the copper absorbed the energy of the laser radiation. The part of the absorbed energy, higher than the ablation threshold of copper, was used to realize material removal, and the part lower than the ablation threshold of copper was reserved in the copper and turned into residual heat, which diffused to the internal copper due to high thermal conductivity. The ablation threshold of graphite was much smaller than that of copper, so the graphite layer absorbed this part of residual heat from the copper, causing larger scope of graphite at the edge to reach its ablation threshold and then to be removed. In the single pulse mode with even frequency, the influence of residual heat diffusion was even more prominent due to high energy peak and long pulse repetition time. In the burst mode, high energy single pulses were split into bursts with small energy peak, to concentrate energy on material removal rather than residual heat diffusion.
Fig. 7. SEM images of kerf edge of graphite anode material in different operating modes. (a) Single pulse mode, (b) burst mode with number of pulses in a burst N = 6, (c) burst mode with number of pulses in a burst N = 10.
The experiment of maximum cutting speed Vm for graphite anode material was carried out at different numbers of pulses in a burst and different laser powers, and the results are shown in Fig. 8. In the adjustable power range of a low power laser, the influence of the increase in the number of pulses in a burst on maximum cutting speed was much more significant than that of the increase in power. Maximum Vm of 500 mm/s was obtained at 37.4 W and number of pulses in a burst of 10, and it was 2.5 times higher than in the single pulse mode. Increasing laser power can raise the maximum cutting speed for anode. It has been estimated in recent research efforts that, to achieve a competitive speed of 500 mm/s that the mechanical cutting can reach, a laser power of 296 W is required if ultrafast laser with pulse width of 10 ps is used [12]. This is too high requirement for a laser. Increasing the number of pulses in a burst can effectively raise the laser cutting speed, and the speed for industrial production can be realized on a low power ps laser.
Fig. 8. Maximum cutting speed for graphite anode material at different numbers of pulses in a burst and different laser powers, and kerf morphology at number of pulses in a burst of 10 and laser power of 37.4 W.
4. Conclusions
In this paper, ps laser cutting of anode material for Li-ion battery in burst mode was introduced, and the influences of the number of pulses in a burst on kerf edge were studied in respect of HAZ and delamination width. The following conclusions were drawn:
- (1) In laser cutting of copper current collector, there was a HAZ with color change at kerf edge, and the HAZ could be effectively reduced by increasing the number of pulses in a burst. At a fixed scanning speed of 600 mm/s, the minimum HAZ of ∼35 µm was obtained at the number of pulses in a burst N = 8. The mechanism of cutting of copper current collector in the burst mode was changed. During cutting, the ps laser pulse irradiation made copper current collector removed directly by rapid thermal gasification and taking away most of the heat. A little residual heat melted the copper, which then resolidified to form a recast layer at the kerf. Under high speed cutting condition, HAZ with color change could be completely eliminated, and EDX analysis showed that the heat influence at the edge was limited within the recast region.
- (2) When the burst mode was used for cutting graphite anode composite material, increasing the number of pulses in a burst could reduce the delamination width. When the number of pulses in a burst N = 10, the delamination width of 35 µm was obtained. The laser cutting of graphite anode with sandwich structure was mainly influenced by the middle copper conductor layer. The burst mode made the heat concentrated on material removal rather than residual heat diffusion, to reduce the delamination width.
- (3) The cutting efficiency for anode material was expressed with maximum cutting speed and was influenced by the number of pulses in a burst and the laser power. The maximum cutting speed increased with the increase in the number of pulses in a burst. The maximum Vm values for copper current collector and graphite anode material obtained at 37.4 W and number of pulses in a burst N = 10 were 3,800 mm/s and 500 mm/s respectively. In addition, the increase in maximum cutting speed caused by the increase in the number of pulses in a burst was evidently larger than that caused by the increase in laser power. For a low power picosecond laser, the power increase range was limited, but the burst mode could realize rapid machining with a low power laser.
Funding
National Natural Science Foundation of China (grant no. 62073089).
Acknowledgments
The authors thank Guansen Pei for his support and help in this study.
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. S. Y. Chung, J. Bloking, and Y. M. Chiang, “Electronically conductive phospho-olivines as lithium storage electrodes,” Nat. Mater. 1(2), 123–128 (2002). [CrossRef]
2. N. J. Dudney, “Solid-state thin-film rechargeable batteries,” Mater. Sci. Eng., B 116(3), 245–249 (2005). [CrossRef]
3. T. Sasaki, Y. Ukyo, and P. Novák, “Memory effect in a lithium-ion battery,” Nat. Mater. 12(6), 569–575 (2013). [CrossRef]
4. Z. Yancheng, W. Chao-Yang, and T. Xidong, “Cycling degradation of an automotive LiFePO4 lithium-ion battery,” J. Power Sources 196(3), 1513–1520 (2011). [CrossRef]
5. S. Li, H. Wang, S. Jack Hu, Y.-T. Lin, and J. A. Abell, “Automatic generation of assembly system configuration with equipment selection for automotive battery manufacturing,” J. Manuf. Syst. 30(4), 188–195 (2011). [CrossRef]
6. M. R. Kronthaler, F. Schloegl, J. Kurfer, R. Wiedenmann, M. F. Zaeh, and G. Reinhart, “Laser cutting in the production of lithium ion cells,” Phys. Procedia 39, 213–224 (2012). [CrossRef]
7. H. Ou, H. Peng, L. Shi, U. Ismat, W. Zhang, and X. Zhang, “Investigation on the blanking properties of thin multi-layer electrode of lithium-ion battery,” J. Phys. Chem. Solids 134, 14–20 (2019). [CrossRef]
8. M. Luetke, V. Franke, A. Techel, T. Himmer, U. Klotzbach, A. Wetzig, and E. Beyer, “A Comparative study on cutting electrodes for batteries with lasers,” Phys. Procedia 12, 286–291 (2011). [CrossRef]
9. D. Lee, R. Patwa, H. Herfurth, and J. Mazumder, “Computational and experimental studies of laser cutting of the current collectors for lithium-ion batteries,” J. Power Sources 210, 327–338 (2012). [CrossRef]
10. D. Lee, R. Patwa, H. Herfurth, and J. Mazumder, “High speed remote laser cutting of electrodes for lithium-ion batteries: anode,” J. Power Sources 240, 368–380 (2013). [CrossRef]
11. A. H. A. Lutey, A. Fortunato, A. Ascari, S. Carmignato, and C. Leone, “Laser cutting of lithium iron phosphate battery electrodes: characterization of process efficiency and quality,” Opt. Laser Technol. 65, 164–174 (2015). [CrossRef]
12. Y. Zhang, J. Li, R. Yang, T. Liu, and Y. Yan, “Analysis of kerf quality on ultrafast laser cutting of anode material for lithium-ion battery,” Opt. Laser Eng. 118, 14–21 (2019). [CrossRef]
13. B. Schmieder, “Analytical model of the laser ablation mechanism of lithium-ion battery coatings,” Proc. SPIE 9351, 93511C (2015). [CrossRef]
14. A. H. A. Lutey, A. Fortunato, S. Carmignato, and M. Fiorini, “High speed pulsed laser cutting of LiCoO2 Li-ion battery electrodes,” Opt. Laser Technol. 94, 90–96 (2017). [CrossRef]
15. W. Pfleging, Y. Zheng, M. Mangang, M. Bruns, and P. Smyrek, “Laser processes and analytics for high power 3D battery materials,” Proc. SPIE 9740, 974013 (2016). [CrossRef]
16. K. Can, K. Hamit, E. Parviz, Ç. Barbaros, K. Denizhan K, A. Önder, Y. Seydi, A. Mehmet D, Ö. Bülent, H. Heinar, H. Ronald, and I. F. Ömer, “Ablation-cooled material removal with ultrafast bursts of pulses,” Nature 537(7618), 84–88 (2016). [CrossRef]
17. G. Giannuzzi, C. Gaudiuso, C. Di Franco, G. Scamarcio, P. M. Lugarà, and A. Ancona, “Large area laser-induced periodic surface structures on steel by bursts of femtosecond pulses with picosecond delays,” Opt. Laser Eng. 114, 15–21 (2019). [CrossRef]
18. G. Caterina, G. Giuseppe, V. Annalisa, L. P. Mario, C. Isabelle, and A. Antonio, “Incubation during laser ablation with bursts of femtosecond pulses with picosecond delays,” Opt. Express 26(4), 3801–3813 (2018). [CrossRef]
19. R. Saeid, L. Jianzhao, and H. Peter R, “Burst train generator of high energy femtosecond laser pulses for driving heat accumulation effect during micromachining,” Opt. Lett. 40(9), 2064–2067 (2015). [CrossRef]
20. M. Domke, V. Matylitsky, and S. Stroj, “Surface ablation efficiency and quality of fs lasers in single-pulse mode, fs lasers in burst mode, and ns lasers,” Appl. Surf. Sci. 505, 144594 (2020). [CrossRef]
21. S. Butkus, V. Jukna, D. Paipulas, M. Barkauskas, and V. Sirutkaitis, “Micromachining of Invar Foils with GHz, MHz and kHz Femtosecond Burst Modes,” Micromachines 11(8), 733 (2020). [CrossRef]
22. D. Metzner, P. Lickschat, and S. Weißmantel, “High-quality surface treatment using GHz burst mode with tunable ultrashort pulses,” Appl. Surf. Sci. 531, 147270 (2020). [CrossRef]
23. D. J. Förster, S. Faas, R. Weber, and T. Graf, “Thrust enhancement and propellant conservation for laser propulsion using ultra-short double pulses,” Appl. Surf. Sci. 510, 145391 (2020). [CrossRef]
