Open Access
Add to BibSonomy
Abstract
Laser sources which produce GHz bursts of ultrashort pulses attract a lot of attention by demonstrating superior performance in material processing. Flexibility of the laser source in a selection of parameters for custom application is highly preferable. In this work, we demonstrate a very versatile method for burst formation using the active fiber loop (AFL). It allows forming GHz bursts containing from 2 up to approximately 2200 pulses in a burst (1000 ns burst width) with identical pulse separation and any predefined intra-burst pulse repetition rate (PRR). The burst pre-shaping by the amplification conditions in the AFL and by the modulation of transmission of the acousto-optic modulator was demonstrated. Industrial-grade ultrafast laser system was able to operate in the single-pulse and GHz-burst regimes. The laser system delivered high-quality 368 fs duration (FWHM) pulses of 15.3 µJ pulse energy and 30.6 W average output power at 2 MHz PRR in the single-pulse regime. In the GHz-burst operation regime, bursts of 2.2 GHz intra-burst repetition rate were formed and amplified to more than 30 W average output power with a burst energy up to 135 µJ at a burst repetition rate of 200 kHz. The sub-picosecond duration of pulses was obtained in the GHz-burst regime at different burst widths.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Discoveries of new light–matter interaction regimes stimulate the development of lasers with increased functionality, and vice versa. Ultrafast lasers are among the most rapidly developed light sources both in terms of parameters and in choice of operation regimes. Femtosecond pulses in the infrared and visible spectral range have found numerous applications in science [1–3], industry [4–9], medicine and biotechnology [8,10–12], and art [13]. Femtosecond and picosecond pulses feature much higher peak power as compared to nanosecond pulses of the same energy. As a result, nonlinear effects are predominant in ultrafast laser processing, allowing to perform machining of transparent materials [4,10,14]. Laser processing, in which removal of the material takes place, involves cutting, milling, scribing, drilling, and surface microstructuring [5,6,15–17]. Among advantages of the femtosecond laser processing are the high precision and small heat-affected zone [4,18,19].
Historically, the first femtosecond lasers applied to material processing were the Ti:Sapphire lasers [4,18]. A high cost, large footprint of the laser system, and low machining throughput have been identified as the main disadvantages of these lasers at the dawn of femtosecond processing. For more than a decade, fiber, hybrid, or rare-earth-doped solid-state lasers have been used as cost-effective alternatives to bulky and expensive Ti:Sapphire laser systems [4]. Nowadays, industrial femtosecond micromachining relies exceptionally on rare-earth-doped (mainly ytterbium-doped) fiber or solid-state lasers [5]. To date, commercially available industrial fiber lasers provide up to 100 W average power femtosecond radiation [20–22]. Hybrid-technology and solid-state lasers with conventional rod-type active media provide >100 W average power femtosecond radiation [5,21]. The majority of them have a master oscillator–power amplifier (MOPA) design. Master oscillators are mostly mode-locked fiber oscillators with pulse repetition rates (PRR) from 50 MHz up to several hundreds of MHz, sometimes even up to GHz PRRs. Chirped pulse amplification (CPA) is employed to avoid nonlinear pulse distortion issues in fiber and solid-state amplifiers.
In parallel with the development of femtosecond lasers, methods to increase the material processing throughput while preserving the benefits of the femtosecond lasers are investigated. The energy of femtosecond pulses can be localized in a small volume; therefore, a small amount of material is removed by a single pulse. A straightforward solution to increase the processing speed or throughput is to increase the number of pulses within a time interval. This was demonstrated in percussion drilling experiments of copper sheets by tens-of-µJ pulses increasing the pulse repetition rate from 50 kHz to about 1 MHz [15]. Drilling of stainless steel, which features lower thermal relaxation time, also manifested an increase in a drill-through speed but with a reduced quality. Trying to understand the light–matter interaction for materials that exhibit free electrons (metals and semiconductors) a two-temperature model was adopted [16,18,23]. Other models which predict the behavior of dielectrics were also developed [24–27]. It is clear that the most suitable set of laser parameters strongly depends on the origin of the material. Even metals differ greatly from each other [15,16,28–31].
When processing the spot by multiple pulses separated by short time intervals from tens of nanoseconds down to tens of picoseconds (corresponding to repetition rates higher than 100 MHz), a local temperature within the processed volume increases. Therefore, subsequent pulses of reduced energy can remove the material, and the average ablation threshold for the sequence of pulses is lower than that for a single pulse [14,25,32,33]. If the repetition rate is high enough that the processed volume does not cool down before the next laser pulse impinges, the so-called cold-ablation takes place [25,34]. The heat is removed with the ablation material, while the surrounding zone remains cold.
In recent years, numerous publications presented results of material processing by femtosecond laser pulses repeating at tens to hundreds of MHz or up to GHz [16,29,35–45]. Various techniques to generate MHz bursts [35,38,46–48] and GHz bursts [20,32,49–52] of pulses were proposed. Although there is contradicting information about whether the GHz repetition rate gives an advantage in terms of ablation throughput and quality, it stimulates a further investigation and optimization of laser parameters [17,36,53]. The newest trend – a burst-in-a-burst regime, which is a combination of MHz and GHz bursts [39,54,55]. High-density bursts have the GHz intra-burst repetition rate, while these bursts also form groups with the internal repetition rate within MHz. These groups of MHz bursts may repeat at rates from 10 Hz to 100 kHz.
It has been theoretically and experimentally proven that the fluence value of individual pulses should be selected carefully. There is an optimal fluence value for which the highest ablation rate, defined as the ablated volume per unit time per unit incident power (mm3 min-1 W-1), is obtained [16,17,36,40,53]. Hodgson et al [16] used the MHz-burst or GHz-burst regime as the means to reduce the fluence of individual pulses. However, by doing this, i.e. by increasing the PRR to reduce the pulse fluence, the two parameters – the time interval between pulses and the fluence – are tied, they cannot be independently varied and the investigation of the burst regimes is not comprehensive. Žemaitis et al [36] optimized the fluence by changing a laser beam diameter, therefore, the fluence could be varied independently from the temporal parameters of the bursts.
For fluence values above the ablation threshold, an important factor is plasma and particle shielding, especially for metals. A number of pulses within a burst plays a significant role [29,40,41,54,56]. Plasma shielding and material re-deposition are material- and fluence- dependent and also distinct for different processes and processing geometries, and even spot sizes [16,40,45,53]. Obata et al [56] demonstrated that in some cases long bursts suffer from plasma shielding. Many papers report dependence on whether an odd or an even number of pulses is used [40,54]. Bonamis et al [53] performed material processing with the fluence of individual pulses far below the ablation threshold, therefore, relatively long bursts were needed to achieve higher ablation efficiency as compared to the single-pulse regime but plasma shielding was avoided. Moreover, it is believed that the shape of the bursts also influences the evolution and final result of the material processing. Therefore, the property of the laser source to shape the bursts is highly preferable. Also, shaping of the bursts provided by the seed laser is often necessary in order to compensate a gain saturation of the main amplifier.
Optimization of the femtosecond laser processing in various burst regimes needs a large amount of data to be collected and analyzed. A deep understanding of dependencies on various laser parameters would guarantee the best results of the processing. In industry, an overall system efficiency in terms of the process speed and watts of power usefully exploited is also evaluated. Usually, an end-user of the laser source can control only a few laser parameters, and some of them cannot be varied independently. The flexibility of the laser source operating regimes gives freedom to the material processing operator to set the most suitable combination of parameters. It is desirable to freely control the following parameters of the burst-mode: (1) an intra-burst PRR, (2) a burst repetition rate (BRR), (3) energy of individual pulses (the amplitude envelope of the burst), and (4) a number of pulses within the burst.
The methods to generate bursts of pulses have some limitations associated with the principle of the method itself. GHz bursts formed by modulating the continuous GHz train of pulses with an acousto-optic modulator (AOM) are limited by the speed of AOM to about 10 ns rise/fall times [16,20,52]. The method using PRR multipliers made of cascaded 50/50 fiber couplers and delay lines requires (N + 1) fiber couplers and N fiber delay lines for the multiplication of the PRR by 2N times [52,53]. The precision of the pulse separations in that method depends on the cutting and splicing precision of the fibers, and the amplitudes setting depends on the splitting ratios of the fiber couplers. The absence of dispersion control may lead to temporal stretching of femtosecond pulses. Finally, the formation of the bursts needs AOM anyway. In the method of burst generation made of multiple birefringent crystals, the maximum number of pulses within a burst is 2N, where N is the number of birefringent crystals [32,48]. Amplitudes are governed by rotation angles of the crystals, while pulse separation precision depends on the thicknesses of the crystals. Interferometric-type and most of the other devices based on beam splitting and delaying a part of radiation have geometrical limitations for making bursts of ultra-high PRR, and, as in previous methods, the burst shape depends on the splitting ratios of the beam splitters [38,45]. Empty resonators with semi-transparent mirrors or etalons [47] can generate a decaying sequence of pulses. Resonators [46,51] with electronically driven optical switches again are limited by ns-time-scale fronts of the switches, and adjustment of amplitudes is not flexible [57].
Previously, we demonstrated [50] an active fiber loop (AFL) – an all-fiber solution to form GHz bursts with identical pulse separation, freely selectable burst width, and controllable shape. The loop-design, loss and dispersion compensation mechanisms allowed to obtain GHz bursts of up to 20 pulses of a few hundred femtosecond pulse duration. After amplification in an Yb-doped fiber power amplifier, 72 nJ burst energy (3.6 nJ energy of individual pulses) at 500 kHz of BRR was achieved. A possibility to form longer bursts containing several thousand pulses was announced. The proposed AFL can be implemented as an individual module or as a part of a burst-mode seeder.
In this work, we demonstrate the formation of 2.2 GHz intra-burst PRR bursts with the AFL and their amplification in an industrial-grade 30 W-level average output power hybrid – an Yb-doped fiber and Yb:YAG – power amplifier. The ultrafast laser was able to operate in the single-pulse and GHz-burst regimes. The presented technique allowed to form bursts containing from 2 up to approximately 2200 pulses. Possibilities of pre-shaping of the bursts at the input of the amplifier allowed achieving the desired burst shape at the output.
2. Experimental results and discussion
2.1 Active fiber loop for GHz burst formation
The active fiber loop (AFL) – the key element for burst formation – depicted in Fig. 1 consists of four main parts: a fiber coupler (FC), ytterbium-doped fiber (YDF) amplifier, chirped fiber Bragg grating (CFBG), and acousto-optic modulator (AOM2). The FC with a splitting ratio of 50/50 has two input and two output ports and forms a loop by connecting one input and one output port (IN2 and OUT2 ports in Fig. 1) with all containing components. All fiber-optic components and optical fibers used in the system are polarization-maintaining and single-mode ensuring a robust all-in-fiber design. The loss and dispersion compensation mechanisms are employed in the AFL. A total loss of the AFL is compensated in the YDF amplifier pumped by a single-mode laser diode (LD). Amplitudes of pulses inside the formed GHz bursts can be adjusted by the amplification conditions as long as amplification is not limited by gain saturation. Dispersion compensation is implemented using the CFBG. The aim is to compensate the chromatic dispersion at each round-trip in the AFL. This is necessary for forming GHz bursts of broadband radiation and attaining a few hundred femtosecond pulse duration at an output of the whole system. The AOM2 inside the AFL is used to actively control a number of pulses within bursts during a burst formation stage. When a desired number of pulses in the burst is formed, the AOM2 is closed (zero transmission of radiation). The formation of the bursts is restarted when the AOM2 is turned on again (the highest transmission of radiation).
Fig. 1. Schematic setup of the active fiber loop with auxiliary AOMs for temporal control of GHz bursts. FC – 2 × 2 fiber coupler (50/50 splitting ratio), CIRC – optical circulator, YDF – ytterbium-doped fiber, CFBG – chirped fiber Bragg grating, LD – single-mode laser diode, AOM1–3 – acousto-optic modulators, PF – a segment of a passive optical fiber. IN1,2 – input ports of the fiber coupler, OUT1,2 – output ports of the fiber coupler. Time delays: T0 – between single input pulses, T1 – between a delayed replica of an input pulse and an undelayed replica of the pulse, T2 – intra-burst pulse separation of the formed bursts, T3 – between bursts of pulses.
The principle of operation of the AFL is based on delaying a part of the radiation inside the AFL and is described in detail in our previous works [50,51]. A desired intra-burst pulse separation T2 of the formed bursts is determined by a delay T1 inside the AFL and a pulse period T0 of an initial pulse train (T2 = T1 – T0). The ultra-high intra-burst PRR is obtained by making the physical length of the loop slightly longer than the cavity length of the master oscillator. The burst formation setup contains the AFL and two auxiliary modulators (AOM1 and AOM3 in Fig. 1) for temporal control of GHz bursts. For the formation of short bursts (shorter than T0), AOM1 is inactive. Pulses of the initial pulse train coupled into the FC are divided into two output ports OUT1 and OUT2 and one part is delayed in the AFL. After each round-trip, the delayed replicas from previous cycles add to an undelayed replica of the next input pulse. Pulses are outcoupled from the AFL through the OUT1 port. A sequence of bursts with an increasing number of pulses is formed at the output of the AFL, while the first pulses of the bursts are separated by the time interval T0. After N round-trips, the last burst in this sequence contains N + 1 pulses. In this case, a sequence of the bursts of pulses containing from 2 to 20 pulses is formed outside the AFL. The AOM3 was used as a pulse picker to select only the bursts with a desired number of pulses determining the burst repetition rate (BRR) or a delay T3 of bursts containing an equal number of pulses (BRR = 1/T3). Experimentally measured 2.2 GHz intra-burst PRR bursts of pulses containing from 2 to 20 pulses in a short-burst formation regime after a different number of round-trips in the AFL are shown in Fig. 2.
Fig. 2. Experimentally measured 2.2 GHz intra-burst PRR burst of pulses containing from 2 to 20 pulses in a short-burst formation regime.
Longer bursts (longer than T0) may be formed using the described method along with a modulation of the initial pulse train using AOM1. The initial stage of the long-burst formation is the same as the formation of short bursts. When pulses that circulate inside the AFL fill the entire time delay between two input pulses T0, the AOM1 blocks the next input pulse. Pulses circulating inside the AFL are outcoupled through the OUT1 port and add up to pulses outcoupled during the previous time interval thus making the burst longer at each cycle. Bursts of different widths can be formed by keeping AOM1 closed and by keeping AOM2 open.
Therefore, the time interval needed for long-burst formation is slightly longer than the time that is needed to fill time delay T0 with pulses. The AOM3 selects the bursts of the desired width and determines the BRR. Experimentally measured 2.2 GHz PRR bursts of 20–1000 ns widths and containing approximately from 40 to 2200 pulses are shown in Fig. 3. The duration of the front and the rear edges of the long bursts were determined by a response time of the AOMs to 10 ns. Furthermore, temporal profiles of the long bursts were affected by the gain saturation effect in the YDF amplifier used in the AFL which resulted in sharp peaks at each 40-pulse portion and reduced amplitude of the trailing pulses in the long-burst formation regime. Smoother shapes of the long bursts may be achieved with additional control of the initial pulse amplitudes by using AOM1 and AOM2, and/or changing the current of the LD which pumps the YDF amplifier. A more sophisticated version of the AFL with a feedback loop and a control algorithm to obtain smoother shapes would be required. It was not performed in this work since the sharply-peaked shape of long bursts may not be crucial for material processing.
Fig. 3. Experimentally measured 2.2 GHz intra-burst PRR burst of pulses containing approx. from 40 to 2200 pulses in a burst (20–1000 ns burst width) in a long-burst formation regime.
One of the main advantages of the technique based on the use of the AFL is the capability to form bursts of laser pulses containing any number of pulses within a burst with identical pulse separation. Furthermore, this technique enables burst formation of any arbitrary intra-burst PRR which do not depend on the initial PRR. To obtain the ultra-high intra-burst PRR, the physical length of the AFL should be slightly longer than the cavity length of the fiber oscillator. The length of the segment of the passive fiber and corresponding time delay T2 predetermine PRR within the formed burst. In the demonstration of >200 GHz PRR within a burst, this segment of passive fiber was chosen to be extremely short only ∼0.95 mm in length. The exact pulse separation within a burst of compressed pulses was examined using a second harmonic noncollinear autocorrelator, as the response time of the conventional semiconductor detectors is too low for time delays of only a few picoseconds. The measured autocorrelation trace of the compressed burst of pulses is shown in Fig. 4. The pulse separation between the pulses within a burst was estimated to be about 4.6 ps which corresponded to a 217 GHz intra-burst PRR. The ultrashort pulse duration of the individual pulses was maintained. The demonstration of 217 GHz intra-burst PRR showed the feasibility and flexibility of this technique and such bursts were not further amplified in the laser system. As demonstrated, this technique has the potential to be used for the development of the laser source which provides bursts of ultrashort laser pulses at THz-level PRR.
Fig. 4. Measured autocorrelation trace of the compressed burst of pulses of 217 GHz PRR. Inset: measured autocorrelation trace with a modified time scale for a clearer representation of the pulse period (T2 = 4.6 ps) between the compressed pulses.
In conclusion, the burst formation technique based on the use of the AFL is a very versatile method as it allows to overcome many limitations encountered by other fiber-based techniques. First of all, short bursts, from 2 pulses in a burst, and long bursts, even up to 1000 ns width, can be formed with identical pulse separation [50,51]. Secondly, any desired intra-burst PRR, even up to THz-level, can be achieved independently from the initial PRR. Moreover, it has the property to adjust the shape of the amplitude envelope of the short bursts by the amplification conditions of the YDF amplifier. Finally, the dispersion compensation in the developed AFL is suitable for the formation of GHz bursts of ultrashort laser pulses.
2.2 Operation of 30 W-level average power ultrashort pulse laser
In this subsection, the operation of an industrial-grade 30 W-level average power femtosecond laser (FemtoLux 30, EKSPLA) with a GHz option is presented. A layout of the ultrafast laser system operating in the single-pulse and GHz-burst regimes is shown in Fig. 5.
Fig. 5. Schematic presentation of the ultrafast 30 W-level average power laser system operating in the single-pulse and GHz-burst regimes. AOM1,3 – acousto-optic modulators, AWG – arbitrary waveform generator.
The laser system consisted of an all-in-fiber seed source, the AFL with auxiliary AOMs for temporal control of GHz bursts, fiber and solid-state power amplifiers, and a pulse compressor. All-in-fiber passively mode-locked seed source operated at 1030.8 nm center wavelength and 50.4 MHz PRR. The seed source pulses were up-chirped to 280 ps duration by using a CFBG temporal stretcher. The CPA configuration of the system was employed to avoid nonlinear pulse distortion issues in the fiber and solid-state power amplifiers. The AFL produced the sequence of bursts with a different number of pulses in the short- and long-burst formation regimes and 2.2 GHz intra-burst PRR as described above. The laser system was able to operate in the single-pulse operation regime (MHz-level PRR) when the AOM2 was closed (zero transmission of radiation).
Burst amplitude envelope pre-shaping possibilities were demonstrated using the AOM3 which was controlled by the arbitrary waveform generator (AWG). The demonstration was performed using a long 500 ns width GHz burst. Various shapes of the burst were achieved using the AOM3 and integrated into the FemtoLux 30 burst pre-shaping layout. A decaying, rising, triangular, and bell-shaped-dipped amplitude envelope of the bursts was obtained by the temporally modulated transmission of the AOM3 (Fig. 6).
Fig. 6. Burst amplitude envelope pre-shaping using the AOM3 controlled by the AWG obtaining decaying, rising, triangular, and bell-shaped-dipped amplitude envelope of the bursts.
Laser pulses in the single-pulse and GHz-burst regimes were amplified to more than 30 W average output power in hybrid Yb-doped fiber / Yb:YAG power amplifiers [58]. The GHz bursts of 1 MHz BRR containing from 2 to 20 pulses of equal amplitudes were amplified up to 31.5 W average output power resulting in 31.5 µJ burst energy (Fig. 7). Burst energy was increased to 138 µJ (27.6 W average output power) by reducing BRR to 200 kHz. The energy of individual pulses in a burst of 2–20 pulses was obtained from 69 µJ to 6.9 µJ, respectively.
Fig. 7. Measured 2.2 GHz intra-burst PRR burst of pulses containing a different number of pulses of equal amplitudes at 31.5 W average output power.
The AFL has the property to form a rising amplitude envelope of a burst as long as it is not limited by the gain saturation, as mentioned above. A rising amplitude envelope of the burst is obtained if the gain is greater than the losses in the AFL. Measured 2.2 GHz intra-burst PRR bursts of pulses containing a different number of pulses of rising amplitudes (first and last pulse amplitude ratio of about 1:2) at 31.5 W average output power are shown in Fig. 8.
Fig. 8. Measured 2.2 GHz intra-burst PRR burst of pulses containing a different number of pulses of rising amplitudes at 31.5 W average output power.
The Yb-doped fiber and Yb:YAG power amplifiers have a high saturation fluence which results in the capability of superior energy storage. However, pulse amplification up to µJ-energy-level leads to saturation-induced pulse envelope deformations of sufficiently long pulses. This effect strongly limits the amplification of long GHz bursts as most of the power is extracted by the front part of the leading pulses. An exponentially decaying burst shape was observed after amplification of the non-pre-shaped long bursts due to the reduction of the inversion for the trailing part of the bursts. GHz non-pre-shaped bursts of 20–500 ns width were amplified at 233–700 kHz BRRs to 32.5 W average output power and up to 140 µJ burst energy. The gain saturation effect was most evident when amplifying long GHz bursts at the lowest BRR of 233 kHz as shown in Fig. 9.
Fig. 9. Measured 2.2 GHz non-pre-shaped bursts of 20–500 ns width at 233 kHz BRR and 32.5 W average output power.
The influence of the gain saturation on the shape of the bursts can be reduced by the burst pre-shaping before amplification in the power amplifiers. An exponentially rising burst shape (as shown in Fig. 6) was performed by the temporally modulated transmission of the AOM3 which was controlled by the AWG. The desired rectangular-like 20–500 ns width GHz burst shapes were achieved at 233 kHz BRR and 31.6 W average output power (135 µJ burst energy) as depicted in Fig. 10.
Fig. 10. Measured 2.2 GHz AOM3 pre-shaped bursts of 20–500 ns width at 233 kHz BRR and 31.6 W average output power for the desired rectangular-like burst shape at the output of the system.
Pulse compression was performed after amplification using a diffraction grating compressor and examined using a second harmonic non-collinear autocorrelator. Autocorrelation functions (ACF) of the compressed pulses in the GHz burst operation regime (bursts containing from 5 to about 1100 pulses (500 ns width)) were compared to the single-pulse regime and depicted in Fig. 11. The measured ACF of the single-pulse regime at 2 MHz PRR and 30.6 W average output power (15.3 µJ pulse energy) had a width of 519 fs (a full-width at half-maximum, FWHM) which corresponded to 368 fs pulse duration for a Gaussian-shaped pulse. Pulse quality degradation was not observed since nonlinear effects did not limit pulse amplification in the power amplifiers. Significant pulse elongation was observed in the GHz-burst regime. The average ACF width, FWHM, was in the range from 605 fs for a 5-pulses-burst to 1369 fs for a 500 ns width burst. The largest increase of the average ACF width was attributed to the narrowing of the pulse spectrum caused by the bell-shaped reflectivity profile of the CFBG used for dispersion compensation in the AFL (the inset of Fig. 12). The long bursts required a high number of round-trips in the AFL which caused the narrowing of the pulse spectrum. Another reason for the difference in the measured ACFs of the compressed pulses in the single-pulse and GHz-burst regimes may be a higher-order dispersion mismatch in the AFL components.
Fig. 11. Measured autocorrelation functions of compressed pulses in the single-pulse and GHz-burst regimes.
Fig. 12. Pulse spectra at the output of the system in the single-pulse and GHz-burst regimes. Inset: reflectivity profile of the CFBG used in the AFL for dispersion compensation.
Pulse spectra at the output of the system in the single-pulse and GHz-burst regimes are compared in Fig. 12. The bell-shaped reflectivity profile of the CFBG used in the AFL was responsible for the major pulse spectrum narrowing, as mentioned above. A Fourier-transform-limited (FTL) pulse duration derived from the measured pulse spectrum increased from 320 fs in the single-pulse regime to 540 fs in the GHz-burst regime for 500 ns burst width. Moreover, a non-uniform reflectivity profile and possible small group delay distortions caused the spectral modulations most clearly visible for the spectrum of the 500 ns width burst. A flat-top CFBG design must be used in the AFL to avoid the pulse spectrum narrowing. Furthermore, a smooth group delay profile of the CFBG and precise dispersion matching would prevent a significant pulse elongation in the GHz-burst regime. Nevertheless, ultrashort pulse duration of sub-1 ps assuming Gaussian-shaped pulse was achieved both in the single-pulse and GHz-burst operation regimes.
3. Conclusion
In this work, a versatile method to form GHz bursts with identical pulse separation, any predefined intra-burst PRR and amplitude envelope using the AFL was demonstrated. It allowed to form 2.2 GHz bursts containing from 2 up to approximately 2200 pulses (1000 ns burst width) in the bursts. The burst pre-shaping by the amplification conditions of the YDF amplifier in the AFL and by the temporally modulated transmission of the AOM resulted in desired burst shapes at the output of the system. The experimentally used EKSPLA FemtoLux 30 laser was able to operate in the single-pulse and GHz-burst regimes. The laser system delivered high-quality 368 fs duration (FWHM) pulses of 15.3 µJ pulse energy and 30.6 W average output power at 2 MHz PRR in the single-pulse regime. In the GHz-burst regime, bursts of 2.2 GHz intra-burst repetition rate were formed and amplified to more than 30 W average output power. The laser system was able to deliver short-burst of GHz pulses at 0.2–1 MHz BRRs and long-bursts at 233–700 kHz. Such BRRs are convenient for most material processing applications since it has a practical upper limit at which various translation and rotation stages used for scanning and beam delivery can be controlled. The highest burst energy of about 140 µJ was achieved at a burst repetition rate of 233 kHz. The ultrashort pulse duration of sub-1 ps assuming Gaussian-shaped pulse was achieved in all GHz-burst regimes for different burst widths despite the evident pulse spectrum narrowing in the AFL. A flat-top CFBG reflectivity profile must be used in the AFL to avoid the pulse spectrum narrowing and to maintain the average pulse duration in the burst not distorted.
Funding
Eurostars (E113855); H2020 (779373).
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. H. Fattahi, Z. Fattahi, and A. Ghorbani, “Prospects of third-generation femtosecond laser technology in biological spectromicroscopy,” J. Opt. 20(5), 054005 (2018). [CrossRef]
2. R. Berera, R. van Grondelle, and J. T. M. Kennis, “Ultrafast transient absorption spectroscopy: principles and application to photosynthetic systems,” Photosynth. Res. 101(2-3), 105–118 (2009). [CrossRef]
3. G. Matthäus, T. Schreiber, J. Limpert, S. Nolte, G. Torosyan, R. Beigang, S. Riehemann, G. Notni, and A. Tünnermann, “Surface-emitted THz generation using a compact ultrashort pulse fiber amplifier at 1060 nm,” Opt. Commun. 261(1), 114–117 (2006). [CrossRef]
4. K. Sugioka and Y. Cheng, “Ultrafast lasers—reliable tools for advanced materials processing,” Light: Sci. Appl. 3(4), e149 (2014). [CrossRef]
5. G. Račiukaitis, “Ultra-Short Pulse Lasers for Microfabrication: A Review,” IEEE J. Sel. Top. Quantum Electron. 27(6), 1–12 (2021). [CrossRef]
6. K. C. Phillips, H. H. Gandhi, E. Mazur, and S. K. Sundaram, “Ultrafast laser processing of materials: a review,” Adv. Opt. Photonics 7(4), 684–712 (2015). [CrossRef]
7. B. Franta, E. Mazur, and S. K. Sundaram, “Ultrafast laser processing of silicon for photovoltaics,” Int. Mater. Rev. 63(4), 227–240 (2018). [CrossRef]
8. M. Malinauskas, A. Žukauskas, S. Hasegawa, Y. Hayasaki, V. Mizeikis, R. Buividas, and S. Juodkazis, “Ultrafast laser processing of materials: from science to industry,” Light: Sci. Appl. 5(8), e16133 (2016). [CrossRef]
9. S. Ahn, J. Kim, D. Lee, C. Park, C. Zander, S.-Y. Ji, and W. S. Chang, “Enhancement of electrical conductivity during the femtosecond laser trimming process for OLED repair,” Opt. Lasers Eng. 137, 106381 (2021). [CrossRef]
10. C. L. Hoy, O. Ferhanoglu, M. Yildirim, K. H. Kim, S. S. Karajanagi, K. M. C. Chan, J. B. Kobler, S. M. Zeitels, and A. Ben-Yakar, “Clinical Ultrafast Laser Surgery: Recent Advances and Future Directions,” IEEE J. Sel. Top. Quantum Electron. 20(2), 242–255 (2014). [CrossRef]
11. S. H. Chung and E. Mazur, “Surgical applications of femtosecond lasers,” J. Biophotonics 2(10), 557–572 (2009). [CrossRef]
12. S. Nagappan, L. Liu, R. Fetcho, J. Nguyen, N. Nishimura, R. E. Radwanski, S. Lieberman, E. Baird-Daniel, H. Ma, M. Zhao, C. B. Schaffer, and T. H. Schwartz, “In Vivo Femtosecond Laser Subsurface Cortical Microtransections Attenuate Acute Rat Focal Seizures,” Cereb. Cortex 29(8), 3415–3426 (2019). [CrossRef]
13. T. Harada, S. Spence, A. Margiolakis, S. Deckoff-Jones, R. Ploeger, A. Shugar, J. Hamm, K. Dani, and A. Dani, “Obtaining Cross-Sections of Paint Layers in Cultural Artifacts Using Femtosecond Pulsed Lasers,” Materials 10(2), 107 (2017). [CrossRef]
14. R. R. Gattass and E. Mazur, “Femtosecond laser micromachining in transparent materials,” Nat. Photonics 2(4), 219–225 (2008). [CrossRef]
15. A. Ancona, F. Röser, K. Rademaker, J. Limpert, S. Nolte, and A. and Tünnermann, “High speed laser drilling of metals using a high repetition rate, high average power ultrafast fiber CPA system,” Opt. Express 16(12), 8958–8968 (2008). [CrossRef]
16. N. Hodgson, H. Allegre, A. Caprara, A. Starodoumov, and S. Bettencourt, “Efficiency of ultrafast laser ablation in burst mode as a function of intra-burst repetition rate and pulse fluence,” Proc. SPIE 11676, 12 (2021). [CrossRef]
17. A. Žemaitis, P. Gečys, M. Barkauskas, G. Račiukaitis, and M. Gedvilas, “Highly-efficient laser ablation of copper by bursts of ultrashort tuneable (fs-ps) pulses,” Sci. Rep. 9(1), 12280 (2019). [CrossRef]
18. B. N. Chichkov, C. Momma, S. Nolte, F. von Alvensleben, and A. Tünnermann, “Femtosecond, picosecond and nanosecond laser ablation of solids,” Appl. Phys. A 63(2), 109–115 (1996). [CrossRef]
19. A. Ostendorf, G. Kamlage, U. Klug, F. Korte, and B. N. Chichkov, “Femtosecond versus picosecond laser ablation,” Proc. SPIE 5713, 1 (2005). [CrossRef]
20. Y. Liu, J. Wu, X. Wen, W. Lin, W. Wang, X. Guan, T. Qiao, Y. Guo, W. Wang, X. Wei, and Z. Yang, “>100 W GHz femtosecond burst mode all-fiber laser system at 1.0 µm,” Opt. Express 28(9), 13414–13422 (2020). [CrossRef]
21. https://www.rp-photonics.com/bg/buy_femtosecond_lasers.html Accessed 24 Jan 2022, seen 2022-08-01.
22. J. Wallace, “Picosecond and femtosecond fiber lasers serve industry and science,” Lasers & Sources, Laser Focus World 15 Sept 2020, 1 (2020).
23. B. Jaeggi, B. Neuenschwander, M. Schmid, M. Muralt, J. Zuercher, and U. Hunziker, “Influence of the Pulse Duration in the ps-Regime on the Ablation Efficiency of Metals,” Phys. Procedia 12, 164–171 (2011). [CrossRef]
24. L. Jiang and H. L. Tsai, “Plasma modeling for ultrashort pulse laser ablation of dielectrics,” J. Appl. Phys. 100(2), 023116 (2006). [CrossRef]
25. C. Kerse, H. Kalaycıoğlu, P. Elahi, B. Çetin, D. K. Kesim, Ö Akçaalan, S. Yavaş, M. D. Aşık, B. Öktem, H. Hoogland, R. Holzwarth, and FÖ Ilday, “Ablation-cooled material removal with ultrafast bursts of pulses,” Nature 537(7618), 84–88 (2016). [CrossRef]
26. O. Dematteo Caulier, K. Mishchik, B. Chimier, S. Skupin, A. Bourgeade, C. Javaux Léger, R. Kling, C. Hönninger, J. Lopez, V. Tikhonchuk, and G. Duchateau, “Femtosecond laser pulse train interaction with dielectric materials,” Appl. Phys. Lett. 107(18), 181110 (2015). [CrossRef]
27. P. Boerner, M. Hajri, N. Ackerl, and K. Wegener, “Experimental and theoretical investigation of ultrashort pulsed laser ablation of diamond,” J. Laser Appl. 31(2), 022202 (2019). [CrossRef]
28. R. Le Harzic, D. Breitling, M. Weikert, S. Sommer, C. Föhl, F. Dausinger, S. Valette, C. Donnet, and E. Audouard, “Ablation comparison with low and high energy densities for Cu and Al with ultra-short laser pulses,” Appl. Phys. A 80(7), 1589–1593 (2005). [CrossRef]
29. T. Hirsiger, M. Gafner, S. Remund, M. V. Chaja, A. Urniezius, S. Butkus, and B. Neuenschwander, “Machining metals and silicon with GHz bursts: Surprising tremendous reduction of the specific removal rate for surface texturing applications,” Proc. SPIE 11267, 27 (2020). [CrossRef]
30. J. Winter, M. Spellauge, J. Hermann, C. Eulenkamp, H. P. Huber, and M. Schmidt, “Ultrashort single-pulse laser ablation of stainless steel, aluminium, copper and its dependence on the pulse duration,” Opt. Express 29(10), 14561–14581 (2021). [CrossRef]
31. J. Lopez, G. Mincuzzi, R. Devillard, Y. Zaouter, C. Hönninger, E. Mottay, and R. Kling, “Ablation efficiency of high average power ultrafast laser,” J. Laser Appl. 27(S2), S28008 (2015). [CrossRef]
32. C. Gaudiuso, G. Giannuzzi, A. Volpe, P. M. Lugarà, I. Choquet, and A. Ancona, “Incubation during laser ablation with bursts of femtosecond pulses with picosecond delays,” Opt. Express 26(4), 3801–3813 (2018). [CrossRef]
33. A. Abdelmalek, Z. Bedrane, E.-H. Amara, B. Sotillo, V. Bharadwaj, R. Ramponi, and S. Eaton, “Ablation of Copper Metal Films by Femtosecond Laser Multipulse Irradiation,” Appl. Sci. 8(10), 1826 (2018). [CrossRef]
34. W. Schulz, U. Eppelt, and R. Poprawe, “Review on laser drilling I. Fundamentals, modeling, and simulation,” J. Laser Appl. 25(1), 012006 (2013). [CrossRef]
35. H. Kalaycioğlu, P. Elahi, Ö Akçaalan, and FÖ Ilday, “High-Repetition-Rate Ultrafast Fiber Lasers for Material Processing,” IEEE J. Sel. Top. Quantum Electron. 24(3), 1–12 (2018). [CrossRef]
36. A. Žemaitis, P. Gečys, G. Račiukaitis, and M. Gedvilas, “Efficient ablation by ultra-short pulse lasers,” Procedia CIRP 94, 962–965 (2020). [CrossRef]
37. P. Elahi, Ö Akçaalan, C. Ertek, K. Eken, FÖ Ilday, and H. Kalaycoğlu, “High-power Yb-based all-fiber laser delivering 300 fs pulses for high-speed ablation-cooled material removal,” Opt. Lett. 43(3), 535–538 (2018). [CrossRef]
38. H. Matsumoto, Z. Lin, and J. Kleinert, “Ultrafast laser ablation of copper with ∼GHz bursts,” Proc. SPIE 10519, 1051902 (2018). [CrossRef]
39. 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]
40. D. J. Förster, B. Jäggi, A. Michalowski, and B. Neuenschwander, “Review on Experimental and Theoretical Investigations of Ultra-Short Pulsed Laser Ablation of Metals with Burst Pulses,” Materials 14(12), 3331 (2021). [CrossRef]
41. S. M. Remund, M. Gafner, M. V. Chaja, A. Urniezius, S. Butkus, and B. Neuenschwander, “Milling applications with GHz burst: Investigations concerning the removal rate and machining quality,” Procedia CIRP 94, 850–855 (2020). [CrossRef]
42. S. Schwarz, S. Rung, C. Esen, and R. Hellmann, “Enhanced ablation efficiency using GHz bursts in micromachining fused silica,” Opt. Lett. 46(2), 282–285 (2021). [CrossRef]
43. F. Caballero-Lucas, K. Obata, and K. Sugioka, “Ultrashort pulse GHz burst mode laser ablation at different wavelengths for surface microfabrication,” Proc. SPIE 11673, 9 (2021). [CrossRef]
44. T. Apostolova, V. Kurylo, and I. Gnilitskyi, “Ultrafast Laser Processing of Diamond Materials: A Review,” Front. Phys. 9, 650280 (2021). [CrossRef]
45. O. Balachninaitė, V. Tamulienė, L. Eičas, and V. Vaičaitis, “Laser micromachining of steel and copper using femtosecond laser pulses in GHz burst mode,” Results Phys. 22, 103847 (2021). [CrossRef]
46. S. Rezaei, J. Li, and P. R. Herman, “Burst train generator of high energy femtosecond laser pulses for driving heat accumulation effect during micromachining,” Opt. Lett. 40(9), 2064–2067 (2015). [CrossRef]
47. A. Okhrimchuk, S. Fedotov, I. Glebov, V. Sigaev, and P. Kazansky, “Single shot laser writing with sub-nanosecond and nanosecond bursts of femtosecond pulses,” Sci. Rep. 7(1), 16563 (2017). [CrossRef]
48. B. Dromey, M. Zepf, M. Landreman, K. O’Keeffe, T. Robinson, and S. M. Hooker, “Generation of a train of ultrashort pulses from a compact birefringent crystal array,” Appl. Opt. 46(22), 5142–5146 (2007). [CrossRef]
49. A. Michailovas and T. Bartulevicius, “Method for generating gigahertz bursts of pulses and laser apparatus thereof,” Int. Patent application publication, WO 2021/059003, A1 (2021).
50. T. Bartulevicius, K. Madeikis, L. Veselis, V. Petrauskiene, and A. Michailovas, “Active fiber loop for synthesizing GHz bursts of equidistant ultrashort pulses,” Opt. Express 28(9), 13059–13067 (2020). [CrossRef]
51. M. Barkauskas, K. Neimontas, and V. Butkus, US Patent application publication, US2020/0067260A1 (2020).
52. C. Kerse, H. Kalaycıoğlu, P. Elahi, Ö Akçaalan, and FÖ Ilday, “3.5-GHz intra-burst repetition rate ultrafast Yb-doped fiber laser,” Opt. Commun. 366, 404–409 (2016). [CrossRef]
53. G. Bonamis, E. Audouard, C. Hönninger, J. Lopez, K. Mishchik, E. Mottay, and I. Manek-Hönninger, “Systematic study of laser ablation with GHz bursts of femtosecond pulses,” Opt. Express 28(19), 27702–27714 (2020). [CrossRef]
54. A. Žemaitis, M. Gaidys, P. Gečys, M. Barkauskas, and M. Gedvilas, “Femtosecond laser ablation by bibursts in the MHz and GHz pulse repetition rates,” Opt. Express 29(5), 7641–7653 (2021). [CrossRef]
55. D. Metzner, P. Lickschat, and S. Weißmantel, “Optimization of the ablation process using ultrashort pulsed laser radiation in different burst modes,” J. Laser Appl. 33(1), 012057 (2021). [CrossRef]
56. K. Obata, F. Caballero-Lucas, and K. Sugioka, “Material Processing at GHz Burst Mode by Femtosecond Laser Ablation,” J. Laser Micro Nanoeng. 16(1), 19–23 (2021). [CrossRef]
57. E. Kažukauskas, S. Butkus, P. Tokarski, V. Jukna, M. Barkauskas, and V. Sirutkaitis, “Micromachining of Transparent Biocompatible Polymers Applied in Medicine Using Bursts of Femtosecond Laser Pulses,” Micromachines 11(12), 1093 (2020). [CrossRef]
58. L. Veselis, T. Bartulevičius, K. Madeikis, and A. Michailovas, “Generation of 40 W, 400 fs pulses at 1 MHz repetition rate from efficient, room temperature Yb:YAG double-pass amplifier seeded by fiber CPA system,” Proc. SPIE 11259, 75 (2020). [CrossRef]
