Open Access
Add to BibSonomy
Abstract
We present a femtosecond, 11.48 GHz intra-burst repetition rate deep UV source at 258 nm based on forth-harmonic generation (FHG) of an electro-optic (EO) comb operating in burst mode. Second-harmonic generation (SHG) of the burst-mode EO comb in LiB3O5 (LBO) leads to 3.7 W average power and 242 fs root-mean-square pulse duration. A second stage of SHG is further performed using two separate β-BaB2O4 (BBO) crystals, delivering deep UV pulses at 523 mW and 294 mW, with estimated pulse durations of half-ps and sub-300 fs, respectively. At divided pulse repetition rates of 5.7 GHz and 2.9 GHz, FHG is also demonstrated, highlighting the potential of flexible repetition rate operation at the GHz level.
© 2020 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
The development of ultrafast lasers has attracted massive attentions in recent years, due to their widespread applications in the industry and great potentials in various aspects of scientific research. The mode-locking technique in laser oscillators has been the most common way of starting ultrashort pulse trains. The pulse repetition rate of such an ultrafast laser is inherently equal to the free spectral range of the cavity, typically varying from tens to hundreds of megahertz for solid-state and fiber lasers. Further increase in the repetition rate to the gigahertz (GHz) level will need very compact design and thus face the limit of the cavity length. Driven by various applications, such as laser ablation [1], ultrafast time-domain spectroscopy [2], electron guns [3] and arbitrary waveform generation [4], high-repetition-rate ultrafast lasers at multiple gigahertz level have been developed based on various principles, including harmonic mode-locking [5], minimizing cavity geometry [6], electro-optic (EO) frequency combs [7] and micro-resonator based frequency combs [8]. Among them, EO combs based on commercially available EO modulators can provide not only a robust all-fiberized configuration without cavity restriction, but also ultrashort pulses with continuously tunable repetition rate at the GHz level [9]. Besides, their output also features a frequency comb structure with the presence of tolerable phase noise from radiofrequency (RF) electronics. Most previous research on EO combs focused on their spectral and temporal features, as well as relevant applications, such as spectral broadening and pulse compression [10], spectroscopy [11], optical communications [12] and precision metrology [13]. The potentials of EO combs for power and repetition rate scaling as fiber lasers, flexible temporal burst generation and efficient frequency conversion to the ultraviolet and mid-infrared regions are still rarely explored.
On the other hand, motivated by the application of electron guns [14], we are especially interested in femtosecond deep ultraviolet (UV) sources with GHz repetition rates, which could be derived from a powerful Yb-fiber-based EO comb via the mature forth-harmonic generation (FHG) technology in bulk nonlinear crystals [15–17]. Such femtosecond GHz pulses with high-energy photons can be used to extract ultrashort electron bunches at the same repetition rate, which can be later accelerated for further purposes. A recent progress in this direction is the demonstration of a 258 nm, picosecond deep UV source pumped by a continuous 10 W EO comb operating at 3 GHz to drive the photoinjector of an S-band accelerator for relativistic electron microscopy with multi-bunch operation [3]. The reported FHG conversion efficiency from 1.03 µm to 258 nm is less than 0.5%, obviously limited by the relatively low peak power at this extreme repetition rate. A further increase in the repetition rate would lead to an even lower pump peak power and insufficient deep UV power. Another similar work is a femtosecond GHz-repetition-rate burst laser at 343 nm [18]. Although the burst operation enabled a high UV power of 20 W for material processing, the repetition rate and UV wavelength do not meet the requirements to reach our objective of driving X-band photoinjectors [19] which could, in turn, benefit a series of potential applications, ranging from generating THz radiation [20] and probing structural dynamics [21] to novel compact accelerators [22].
In this paper, we present a femtosecond, 11.48 GHz EO comb at 1.03 µm, operating in burst mode (with a burst repetition rate of 115 kHz) with ∼500 pulses per burst and delivering up to 10 W of average power. Single-pass, two-stage FHG of the burst-mode EO comb is further implemented for deep UV generation. Compared to continuous pulse operation, the burst mode significantly increases pulse peak power and thus the overall FHG conversion efficiency. Besides, it also meets the requirements for our target application of photoinjectors. Using one LiB3O5 (LBO) and two separate β-BaB2O4 (BBO) crystals as nonlinear media, 523 mW (4.5 µJ per burst and 9.1 nJ per pulse) and 294 mW (2.56 µJ per burst and 5.1 nJ per pulse) of deep UV powers are obtained, corresponding to FHG conversion efficiencies of 5% and 2.9%, with estimated pulse durations of half-ps and sub-300 fs (root-mean-square width), respectively. Besides, using frequency division and pulse picking technique, lower intra-burst pulse repetition rates of ∼5.7 GHz and ∼2.9 GHz are also demonstrated at 1.03 µm and further extended to 258 nm. To our knowledge, this source represents the first femtosecond, multi-GHz intra-burst repetition rate laser in the deep UV region.
2. Experimental setup
The experimental setup for the femtosecond burst-mode EO comb and its FHG is shown in Fig. 1. Our EO comb has been described and characterized elsewhere [9]. It originates from a continuous-wave single-frequency diode laser at 1030 nm and comprises two phase modulators for spectral broadening and one intensity modulator for pulse formation. By adjusting the phase delay of the driving wave for the intensity modulator, a nearly linear and negative chirp across each pulse is selected in our experiment. Although the EO comb can provide tunable repetition rate across 11–18 GHz, we fix it at 11.48 GHz, targeting the frequency of X-band photoinjectors [23]. When the laser system is integrated in the photogun, the modulators are directly driven by the accelerator RF clock, enabling straightforward and low timing jitter synchronization.
Fig. 1. Schematic of the femtosecond burst-mode EO comb and its FHG. LD: laser diode, DCF: double-clad fiber, ISO: isolator, AOM: acousto-optic modulator, SMF: single-mode fiber, WDM: wavelength division multiplex, λ/2: half-wave plate, L: lens, DM: dichroic mirror.
The comb light with an average power of 60 mW is first sent into a cladding-pumped amplifier (Nufern, PLMA-YDF-10/125-M, 1.7 m) and amplified to 1 W. For demonstrating various intra-burst pulse repetition rates (below 11 GHz), we later introduced an additional Mach-Zehnder intensity modulator and its driving electronics between the EO comb seed and the fiber amplifier for frequency division and pulse picking. Then, a fiber-coupled acousto-optic modulator (AOM) is used to carve bursts of pulses by optical gating. By setting the burst repetition rate at 115 kHz and the burst duration at 44 ns, we could isolate around 500 pulses at 11.48 GHz repetition rate in each burst, and obtain a reduced pulse number per second of 57.4 million, corresponding to a duty cycle of 0.5%. The output burst-mode EO comb from the AOM is sent into 350 m of passive fiber (Nufern, PM980-XP) for pulse compression. A total fiber normal dispersion of ∼8.75 ps2 compensates the negative linear chirp of the EO comb pulses, and thus compresses the modulated pulses to sub-2 ps. Then, we boost the power of the compressed pulses using a short core-pumped fiber amplifier (CorActive, Yb 401-PM, 0.6 m) to 130 mW and propagate the beam through 4.5 m of passive fiber (Nufern, PM980-XP) for nonlinear spectral broadening and further pulse shortening. The balance between self-phase modulation (SPM) and normal dispersion in the fiber broadens the spectrum and imposes linearized chirp on the pulses, allowing to generate femtosecond pulses with proper compression. In order to further increase the average power, we use a main amplifier consisting of two successive rod-type fiber stages. Following this main amplifier, a grating compressor with ∼75% efficiency is employed to compress the pulses for characterization and further frequency conversion.
In the frequency conversion part, we first pass the amplified burst-mode EO comb through a half-wave plate at 1030 nm to control the beam polarization and then focus the beam into a LBO crystal for frequency doubling, using a lens (L1) of focal length, f=60 mm. The beam waist radius is ∼32 µm, corresponding to a focusing parameter of ξ∼1.0 [24]. The LBO crystal is 10 mm long with a cross section of 3×3 mm2, and is cut at θ=90° (φ=0°) with both end faces antireflection (AR)-coated at 1030 nm and 515 nm. The crystal is mounted in an oven and maintained at 189°C to reach type-I (oo→e) noncritical phase-matching condition for SHG. The generated green beam is separated from the residual pump using a dichroic mirror and later collimated using a lens (L2) of focal length, f=150 mm. For FHG at 258 nm, we further focus the burst-mode green light into two separate BBO crystals with different lengths. A beam waist radius of ∼38 µm is set using a lens (L3) of focal length, f=60 mm, leading to ξ∼0.034. A half-wave plate at 515 nm is also used to adjust the beam polarization direction. The two BBO crystals are 1 mm and 0.2 mm thick with the same cross section of 5×5 mm2. They are cut at θ=50° (φ=0°) with both end faces AR-coated at 515 nm and 258 nm and maintained at room temperature for type-I (oo→e) phase-matching. Then, a harmonic separation mirror which is highly reflecting at 258 nm and AR-coated at 515 nm & 1030 nm is used to extract the generated deep UV source for further characterization.
3. Experimental results
We first characterized the spectrum and pulse of the burst-mode EO comb before nonlinear spectral broadening. Figure 2(a) shows the measured EO comb spectrum at the output of the AOM. The modulated spectrum has a 10 dB bandwidth of ∼2.5 nm, corresponding to 62 comb lines. The spectrum is slightly tilted due to the limited gain bandwidth of the used Yb-doped fiber amplifiers, which is centered at longer wavelength. For measuring the autocorrelation (AC) trace of the compressed pulses with our autocorrelator, we increased the power of the burst-mode EO comb to ∼40 mW at the output of the core-pumped fiber amplifier and obtained a typical EO comb AC profile with a pedestal, having a FWHM duration of 2.5 ps, as shown in Fig. 2(b). Considering a deconvolution factor of 1.4 (typical for EO comb pulses), the compressed pulse duration is 1.78 ps, consistent with our previous results [9], where a grating compressor (introducing third order dispersion) is used instead of a single-mode fiber.
Fig. 2. (a) Measured spectrum and (b) intensity autocorrelation of the burst-mode EO comb at 1030 nm after pulse compression using 350 m of fiber.
On a larger timescale, using an oscilloscope (Keysight, InfiniiVision MSO-X 4154A) with a bandwidth of 1.5 GHz and a fast InGaAs photodetector (Optilab, PD-40x) with a bandwidth of 40 GHz, we also characterized the burst-mode operation of the EO comb. We drove the AOM with a 115 kHz square wave signal, corresponding to a period of ∼8.7 µs between two adjacent bursts, as shown in Fig. 3(a). Figure 3(b) shows a zoomed-in single burst envelop as the pulses are not resolved by the oscilloscope. Compared to the driving square wave signal, the leading and trailing edges of the burst are less steep, which is attributed to the limited response time of the AOM. Using a sampling oscilloscope (Keysight, DCA-X 86100D) with a bandwidth of 45 GHz and the same photodetector, at the output of the EO comb seed, we were also able to measure and confirm the 11.48 GHz pulse train, as shown in Fig. 3(c).
Fig. 3. (a) Burst train at 115 kHz, (b) single optical burst profile (red), RF driving signal (grey) and (c) 11.48 GHz intra-burst pulse train of the EO comb.
Furthermore, for demonstrating the potential of variable pulse repetition rate operation of our burst-mode EO comb system at the GHz level, we also introduced an intensity modulator (iXblue, NIR-MX-LN-20) for repetition rate division. The 11.48 GHz signal from the RF synthesizer is divided by 2 or 4 using a division card to reduce the signal frequency to ∼5.7 GHz and ∼2.9 GHz. The frequency-reduced signal is further sent to a home-made pulse generation card to generate RF pulses with a duration of ∼30 ps. The modulator, biased at “NULL” point by using a DC voltage, is driven by the RF pulses at 5.7 GHz and 2.9 GHz to provide an optical pulse picker. The pulse picking is optimized by adjusting the RF signal delay using a phase shifter. Figures 4(a) and 4(c) show the measured optical pulses picked by the intensity modulator at divided repetition rates. Figures 4(b) and 4(d) are the corresponding measured single burst profiles, where we keep the burst repetition rate at 115 kHz and set the duration two and four times longer than 44 ns in Fig. 3(b), respectively, to always maintain ∼500 pulses per burst and thus the same pulse peak power for later nonlinear spectral broadening. With other RF frequency dividers allowing additional integer number division, we would be able to obtain a wider range of discrete intra-burst repetition rates from sub-1 to 5 GHz. Besides, since our EO comb seed operates over the frequency band of 11–18 GHz [9], fundamental and divided intra-burst repetition rates up to 18 GHz and 9 GHz, respectively, are also realistic, just with the need of using proper fiber length for pulse compression later.
Fig. 4. Measured pulse trains at divided repetition rates of (a) 5.7 GHz and (c) 2.9 GHz, and (b,d) corresponding single burst profiles.
At the fundamental repetition rate of 11.48 GHz, we further studied the spectral broadening and pulse compression part of our burst-mode EO comb system. As the compressed energetic pulses propagate through 4.5 m of fiber before injection to the rod-fiber amplifier, the input spectrum will be significantly broadened through SPM effect, and the pulses will be linearly chirped to a few picoseconds due to SPM together with normal dispersion. For input pulses with Gaussian profile, this process will reshape the pulses towards a parabolic shape which would benefit further high-power amplification [25,26]. In our case, the “M-shaped” EO comb spectrum limited the parabolic shaping, yet this process enabled a broadened output spectrum and compressible pulses. Figure 5(a) shows the measured spectra at the output of the rod-fiber amplifier with two different power levels, 150 mW and 14 W before compression. In both cases, we obtained spectra broad enough to support femtosecond pulse generation. When the power was increased from 150 mW to 14 W, we observed additional modulations on the far spectral wings. We attribute this to optical wave breaking effect, as indicated by the sidelobes of the corresponding AC trace, which sets in through the earlier nonlinear spectral broadening process and becomes more significant during high power amplification. However, the spectrum at 14 W still shows a narrower 10 dB bandwidth than the one at 150 mW, implying that the gain narrowing effect prevails over SPM in our rod-fiber amplifier. The inset shows a zoomed-in measurement at 14 W, where the 11.48 GHz comb lines are also resolved. Figures 5(b) and 5(c) are the corresponding intensity AC measurements at 150 mW and 14 W, as well as the simulated AC traces of transform-limited pulses based on the measured spectra for comparison. The AC FWHM duration is measured to be 456 fs at 150 mW. However, it increased to 642 fs with the appearance of stronger lobes on the far pulse wings for 14 W output power. This can be explained by the gain narrowing effect which squeezes the spectrum and makes it squarer, and optical wave breaking effect with nonlinear chirps on the far wings of the pulse.
Fig. 5. (a) Measured spectra of the burst-mode EO comb at 150 mW and 14 W. Inset: Zoomed-in spectrum at 14 W showing the comb teeth. Intensity autocorrelations of the burst-mode EO comb pulses at (b) 150 mW and (c) 14 W, and the corresponding calculated autocorrelations of the transform-limited (TL) pulses.
Although we used up to 14 W of the burst-mode EO comb for FHG, considering the compressor efficiency of ∼75%, the available power after the compressor is 10.4 W. For characterizing the SHG, we initially performed power scaling measurement for the generated green light, as shown in Fig. 6(a). A 10-mm-long LBO crystal is chosen as the nonlinear medium, with the assistance of SNLO software [27], to ensure the corresponding group velocity mismatch (GVM) of ∼50 fs/mm between the pump and SHG beam does not significantly distort the green pulses. Instead of the expected quadratic relationship, the green power shows a linear dependence on the pump power with a slope efficiency of 37.3%, which, together with the conversion efficiency curve, indicates that the SHG process reaches saturation with high pump depletion. For an input pump power of 10.2 W in front of the LBO crystal, we obtained a maximum green power of 3.7 W. At this maximum power level, we also measured the green spectrum which has a multi-spike profile, as shown in Fig. 6(b). By comparing to the pump spectrum at 14 W, the observed multiple spectral spikes can be explained by SHG of the spectral spikes on the pump spectrum, sum-frequency generation between different spectral spikes on the pump spectrum and back-conversion effect. The corresponding beam profile of the green beam was also captured and shown in the inset. We further performed M2 measurement for this green beam and obtained the beam quality values in the horizontal plane and vertical plane, M2x=2.1, M2y=1.5. The simulated AC trace of the green pulses based on the measured spectrum is also performed, as plotted in Fig. 6(c). Using a home-made autocorrelator comprising a translation stage driven by a motorized actuator, we measured the intensity AC trace of the green pulses, having a FWHM duration of 413 fs, as also shown in Fig. 6(c). For this non-Gaussian intensity profile, we further calculated the root-mean-square (RMS) duration of the AC trace, which is 342 fs, and thus got a RMS pulse duration of 242 fs [28]. In contrast to the AC trace recorded at 1030 nm where satellite pulses are clearly visible, the AC trace for green pulses decays regularly for longer delays. This means the SHG process also works as a pulse filter here, which efficiently suppresses the low-intensity lobes on the pump pulses and leads to cleaner green pulses.
Fig. 6. (a) Variations of the green power and conversion efficiency as functions of the input pump power. (b) Measured spectrum of the generated green light at 3.7 W. Inset: Beam profile of the green beam at the same power. (c) Intensity AC of green pulses measured at 3.7 W and simulated AC of transform-limited pulses.
At last, using this green beam with 11.48 GHz intra-burst repetition rate, we implemented another stage of SHG for deep UV generation. For an input green power of 3.5 W, we generated a maximum power of 523 mW at 258 nm from a 1 mm BBO crystal, corresponding to a SHG conversion efficiency of 14% from 515 nm to 258 nm and an FHG conversion efficiency of 5% from 1030 nm to 258 nm. The obtained deep UV energy is 4.5 µJ/burst and 9.1 nJ/ pulse. Figure 7(a) shows the UV spectrum at full power, measured using a UV spectrometer based on a linear CCD array (Avantes, AvaSpec-ULS3648). For characterizing the stability of the deep UV power, we performed a 10 minutes’ measurement, where the deep UV source exhibited a power instability of 1.6% RMS, as shown in Fig. 7(b). In a longer term of deep UV power monitoring, we observed that the power was gradually declining and linearly dropped to 425 mW after 1 hour’s operation. This power decline is due to inevitable thermal effects caused by linear and two-photon absorption, as well as dynamic color center formation in BBO [17], which limit high-power and long-term operations. For applications where a better long-term power stability is required, a mitigation method could be heating up the BBO crystal and operate at a higher temperature, as demonstrated in [17].
Fig. 7. (a) Measured deep UV spectrum at full power. (b) Power stability measurements with 1 mm and 0.2 mm BBO crystals.
Considering the GVM between the green and deep UV pulses in the BBO crystal is ∼630 fs/mm, and the green RMS pulse duration of 242 fs, the 1 mm BBO crystal will prolong deep UV pulses to ∼500 fs, estimated by SNLO software. As an attempt to maintain a short pulse duration, we also used a 0.2 mm thick BBO crystal as the nonlinear medium, which has almost negligible distortion to the deep UV pulses and thus allows much shorter deep UV pulse emission at sub-300 fs level (RMS width). The obtained maximum power using this 0.2 mm thick BBO is 294 mW (2.56 µJ/burst and 5.1 nJ/pulse), with an instability of 1.9% RMS for 10 minutes, as also plotted in Fig. 7(b). These obtained deep UV powers are well sufficient for our target application of X-band photoinjectors. Besides, using the 1 mm BBO crystal, we further demonstrated the FHG at divided intra-burst repetition rates. Although we were not able to measure the repetitive pulses at GHz level, due to the lack of fast detectors in the UV range, we still recorded deep UV powers of 395 mW at 5.7 GHz and 480 mW at 2.9 GHz, showing the potential of our system for variable GHz intra-burst repetition rate operation.
4. Conclusions
In summary, we have demonstrated a femtosecond, 11.48 GHz intra-burst repetition rate deep UV source at 258 nm based on single-pass two-stage FHG. The pump laser for the FHG originates from an EO comb which experiences acousto-optic modulation for burst-mode operation at 115 kHz with ∼500 pulses per burst, pulse compression and nonlinear spectral broadening in single-mode fibers, power amplification and finally pulse compression in free space. For a pump average power of 10.4 W, we obtained burst-mode green light at 515 nm with an average power of 3.7 W and a calculated RMS pulse duration of 242 fs, through SHG in a 10 mm long LBO crystal. 523 mW of deep UV source at 258 nm, corresponding to 4.5 µJ/burst and 9.1 nJ/ pulse, is further achieved using a 1 mm thick BBO crystal for SHG. The pulse duration is estimated to be half picosecond due to GVM. With a thinner BBO crystal of 0.2 mm, we also obtained 294 mW of deep UV light (2.56 µJ/burst and 5.1 nJ/pulse) with much less pulse distortion caused by GVM, leading to short pulses of sub-300 fs (RMS width). For demonstrating the versatility of our system for various intra-burst repetition rate operation at the GHz level, we also performed FHG at divided repetition rates of 5.7 GHz and 2.9 GHz, where 395 mW and 480 mW of deep UV powers were recorded, respectively. This source represents the first femtosecond, multi-GHz intra-burst repetition rate laser in the deep UV region, meeting the requirements for the application of X-band photoinjectors. Besides, its potential of power scaling and flexible intra-burst repetition rate switching at the GHz level could also benefit a wide range of scientific and industrial applications.
Funding
Horizon 2020 Framework Programme (825246); Agence Nationale de la Recherche (ANR-10-IDEX-03-02); Conseil Régional Aquitaine (2019-1R5M04); Institut Universitaire de France (IUF); Aquitaine Science Transfert (AST_AT_2018-043).
Acknowledgments
The authors acknowledge financial support from the French National Research Agency (ANR) in the frame of “the investments for the future” Programme IdEx Bordeaux – LAPHIA (ANR-10-IDEX-03-02). This project is an initiative of the Photonics Public Private Partnership from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 825246. The authors are also grateful to Julien Didierjean for his assistance with the rod-type fiber amplifier.
Disclosures
The authors declare no conflicts of interest.
References
1. K. Mishchik, G. Bonamis, J. Qiao, J. Lopez, E. Audouard, E. Mottay, C. Hönninger, and I. Manek-Hönninger, “High-efficiency femtosecond ablation of silicon with GHz repetition rate laser source,” Opt. Lett. 44(9), 2193–2196 (2019). [CrossRef]
2. O. Kliebisch, D. C. Heinecke, and T. Dekorsy, “Ultrafast time-domain spectroscopy system using 10 GHz asynchronous optical sampling with 100 kHz scan rate,” Opt. Express 24(26), 29930–29940 (2016). [CrossRef]
3. C. Li, H. Xuan, L. Winkelmann, and I. Hartl, “3 GHz, 257 nm picosecond source for electron guns,” presented at the 8th EPS-QEOD Europhoton Conference, Barcelona, Spain, 2–7 Sept. 2018.
4. S. T. Cundiff and A. M. Weiner, “Optical arbitrary waveform generation,” Nat. Photonics 4(11), 760–766 (2010). [CrossRef]
5. J. Koo, J. Park, J. Lee, Y. M. Jhon, and J. H. Lee, “Femtosecond harmonic mode-locking of a fiber laser at 3.27 GHz using a bulk-like, MoSe2-based saturable absorber,” Opt. Express 24(10), 10575–10589 (2016). [CrossRef]
6. S. Kimura, S. Tani, and Y. Kobayashi, “Kerr-lens mode locking above a 20 GHz repetition rate,” Optica 6(5), 532–533 (2019). [CrossRef]
7. A. J. Metcalf, C. D. Fredrick, R. C. Terrien, S. B. Papp, and S. A. Diddams, “30 GHz electro-optic frequency comb spanning 300 THz in the near infrared and visible,” Opt. Lett. 44(11), 2673–2676 (2019). [CrossRef]
8. A. S. Raja, J. Liu, N. Volet, R. N. Wang, J. He, E. Lucas, R. Bouchandand, P. Morton, J. Bowers, and T. J. Kippenberg, “Chip-based soliton microcomb module using a hybrid semiconductor laser,” Opt. Express 28(3), 2714–2721 (2020). [CrossRef]
9. A. Aubourg, J. Lhermite, S. Hocquet, E. Cormier, and G. Santarelli, “Generation of picosecond laser pulses at 1030 nm with gigahertz range continuously tunable repetition rate,” Opt. Lett. 40(23), 5610–5613 (2015). [CrossRef]
10. D. R. Carlson, D. D. Hickstein, W. Zhang, A. J. Metcalf, F. Quinlan, S. A. Diddams, and S. B. Papp, “Ultrafast electro-optic light with subcycle control,” Science 361(6409), 1358–1363 (2018). [CrossRef]
11. K. Fdil, V. Michaud-Belleau, N. B. Hébert, P. Guay, A. J. Fleisher, J. Deschênes, and J. Genest, “Dual electro-optic frequency comb spectroscopy using pseudo-random modulation,” Opt. Lett. 44(17), 4415–4418 (2019). [CrossRef]
12. V. Torres-Company and A. M. Weiner, ““Optical frequency comb technology for ultra-broadband radio-frequency photonics”,” Laser Photonics Rev. 8(3), 368–393 (2014). [CrossRef]
13. E. L. Teleanu, V. Durán, and V. Torres-Company, “Electro-optic dual-comb interferometer for high-speed vibrometry,” Opt. Express 25(14), 16427–16436 (2017). [CrossRef]
14. D. Vouagner and J. P. Girardeau-Montaut, “Picosecond laser-induced ultrashort electron emission from carbon-based photocathodes,” Thin Solid Films 453-454, 22–26 (2004). [CrossRef]
15. M. Müller, A. Klenke, T. Gottschall, R. Klas, C. Rothhardt, S. Demmler, J. Rothhardt, J. Limpert, and A. Tünnermann, “High-average-power femtosecond laser at 258 nm,” Opt. Lett. 42(14), 2826–2829 (2017). [CrossRef]
16. A. S. Rao, N. A. Chaitanya, and G. K. Samanta, “High-power, high repetition-rate, ultrafast fibre laser based source of DUV radiation at 266 nm,” OSA Continuum 2(1), 99–106 (2019). [CrossRef]
17. S. C. Kumar, J. C. Casals, J. Wei, and M. Ebrahim-Zadeh, “High-power, high-repetition-rate performance characteristics of β-BaB2O4 for single-pass picosecond ultraviolet generation at 266 nm,” Opt. Express 23(21), 28091–28103 (2015). [CrossRef]
18. G. Bonamis, J. Sanabria, E. Audouard, J. Lopez, I. Manek-Hönninger, E. Mottay, and C. Hönninger, “20-W ultraviolet femtosecond GHz burst laser,” in 2019 Conference on Lasers and Electro-Optics Europe and European Quantum Electronics Conference, OSA Technical Digest (Optical Society of America, 2019), paper cf_8_6.
19. R. A. Marsh, G. G. Anderson, S. G. Anderson, and D. J. Gibson, “Performance of a second generation X-band rf photoinjector,” Phys. Rev. Accel. Beams 21(7), 073401 (2018). [CrossRef]
20. Y. Shen, X. Yang, G. L. Carr, Y. Hidaka, J. B. Murphy, and X. Wang, “Tunable few-cycle and multicycle coherent terahertz radiation from relativistic electrons,” Phys. Rev. Lett. 107(20), 204801 (2011). [CrossRef]
21. P. Musumeci, J. T. Moody, C. M. Scoby, M. S. Gutierrez, and M. Westfall, “Laser-induced melting of a single crystal gold sample by time-resolved ultrafast relativistic electron diffraction,” Appl. Phys. Lett. 97(6), 063502 (2010). [CrossRef]
22. X. F. D. Stragier, O. J. Luiten, S. B. van der Geer, M. J. van der Wiel, and G. J. H. Brussaard, “Experimental validation of a radio frequency photogun as external electron injector for a laser wakefield accelerator,” J. Appl. Phys. 110(2), 024910 (2011). [CrossRef]
23. M. A. Prantil, E. Cormier, J. W. Dawson, D. J. Gibson, M. J. Messerly, and C. P. J. Barty, “Widely tunable 11 GHz femtosecond fiber laser based on a nonmode-locked source,” Opt. Lett. 38(17), 3216–3218 (2013). [CrossRef]
24. G. D. Boyd and D. A. Kleinman, “Parametric interaction of focused Gaussian light beams,” J. Appl. Phys. 39(8), 3597–3639 (1968). [CrossRef]
25. W. Fu, Y. Tang, T. S. McComb, T. L. Lowder, and F. W. Wise, “Limits of femtosecond fiber amplification by parabolic pre-shaping,” J. Opt. Soc. Am. B 34(3), A37–A42 (2017). [CrossRef]
26. W. Fu, L. G. Wright, and F. W. Wise, “High-power femtosecond pulses without a modelocked laser,” Optica 4(7), 831–834 (2017). [CrossRef]
27. A. V. Smith, SNLO software, http://www.as-photonics.com/products/snlo.
28. E. Sorokin, G. Tempea, and T. Brabec, “Measurement of the root-mean-square width and the root-mean-square chirp in ultrafast optics,” J. Opt. Soc. Am. B 17(1), 146–150 (2000). [CrossRef]
