Open Access
Add to BibSonomy
Abstract
We report on dispersion management in mid-IR optical parametric chirped pulse amplifiers (OPCPA) aiming for high-energy few-cycle pulses beyond 4 µm. The available pulse shapers in this spectral region limit the feasibility of sufficient higher-order phase control. Intending the generation of high energy pulses at 12 µm via DFG driven by the signal and idler pulses of a midwave-IR OPCPA, we introduce alternative approaches for mid-IR pulse shaping, namely a germanium-prism pair and a sapphire-prism Martinez compressor. Furthermore, we explore the limits of bulk compression in Si and Ge for multi-mJ pulse energies.
© 2023 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
High-energy, few-cycle pulses at kilohertz repetition rate in the midwave-IR (MWIR, 3–8 µm) and longwave-IR (LWIR, 8 - 15 µm) is the prerequisite for fundamental studies of strong-field laser-matter interactions and the non-equilibrium properties of condensed matter [1,2]. For wavelengths beyond 10 µm, the study of molecular vibrations and/or for performing time resolved two-dimensional infrared spectroscopy is of high interest [3].
Due to the lack of directly pumped solid-state lasers beyond 4 µm wavelength, optical parametric chirped pulse amplification (OPCPA) has established as the key technique to generate high-energy few-cycle pulses at kilohertz repetition rates [4–6]. Using such sources as drivers for generating high harmonics [7–9] or X-ray pulses [10], structures and dynamics of materials can be studied with enhanced temporal and spatial resolution.
Spectral phase control is the key to reach the few-cycle regime. For this purpose, pulse shaper designs containing active optical elements such as acousto-optic programmable dispersive filters (AOPDF) or spatial light modulators (SLM) have been implemented [11]. These devices enable the necessary phase control and are well established in the visible (0.4–0.8 µm), near-IR (NIR, 0.8–1.5 µm) and shortwave-IR (SWIR, 1.5–3 µm) spectral range [12–16]. They have also proven their suitability in the mid-IR, as evidenced by the demonstrated few-cycle mid-IR OPCPAs [6].
The challenge for OPCPAs beyond 4 µm is the limited availability of nonlinear crystals being transparent for the pump and the idler pulses and, exhibiting high damage threshold [17]. The unfavorable pump/signal to idler photon energy relation when using near-infrared pump sources (typically around 1 µm) limits the energy scaling for longer idler wavelengths [17]. 2-µm pump lasers are more advantageous and allow exploiting the high nonlinearity of ZnGeP2 (ZGP) crystals for parametric amplification [18]. The availability of such pump sources based on few-ps 2-µm Ho-doped lasers [19,20] paved a new way towards mJ-level few-cycle MWIR and LWIR OPCPAs. This is confirmed by the few demonstrated broadband high-energy OPCPAs beyond 4 µm which rely on 2-µm pump sources [20–22]. The first OPCPA system reaching sub-millijoule pulses beyond 4 µm produced 75 fs at 5.1 µm was demonstrated in 2017 and operated at 1 kHz repetition rate [21]. This system containing an SLM for spectral shaping was further improved with an upgrade to four optical parametric amplifier (OPA) stages and delivered 89 fs with 3.4 mJ pulse energy centered at 4.9 µm wavelength [22]. The corresponding peak power of 33 GW represents the record for OPCPAs beyond 4 µm wavelength. Recently, the front-end concept of this MWIR OPCPA scheme was changed to a fs Cr:ZnS master oscillator. Pumped by a 1 kHz Ho:YLF regenerative amplifier, a two-stage OPCPA containing ZGP crystals delivered tunable idler pulses between 5.4 and 6.8 µm with >0.4 mJ energy. The dispersion management comprised an AOPDF for pulse shaping and enabled compression of the pulses to a sub-100 fs duration [23]. At a longer central wavelength of 7 µm, an OPCPA equipped with five ZGP OPA stages was demonstrated [20]. It delivered 188 fs pulses with energy of 0.75 mJ pumped by a cryogenically-cooled 2 µm Ho:YLF source, however, at a lower repetition rate of 0.1 kHz. The system did not comprise a spectral phase shaper and the peak power amounts to 3.7 GW.
The generation of femtosecond pulses with high energy at wavelengths beyond 10 µm is even more challenging and has been lacking great improvement over the last two decades. Pulse energies on the order of 1 µJ with a sub-200 fs pulse duration have been demonstrated with HgGa2S4 and AgGaS2 [24], AgGaSe2 [25], and GaSe [26–28] nonlinear crystals. Recently, pulse energies of up to 12 µJ beyond 10 µm were achieved via DFG in GaSe [25,29]. With the very recent demonstration of the first LWIR OPCPA new records were set. The three stage OPCPA based on GaSe delivered 65 µJ pulses at 11.4 µm with a pulse duration of 185 fs. This LWIR OPCPA is also pumped at 2 µm and operates at 1 kHz. [30].
There are some special issues when operating OPCPAs in the mid-IR compared to the NIR spectral range, in particular when aiming at high pulse energies. The most promising way is managing the required dispersion for stretching and compression of the pulses exclusively with bulk material. This holds true for the signal as well as for the idler by profiting from the second-order chirp reversal between signal and idler in OPCPAs.
However, the amount of third-order dispersion (TOD), which is positive for all relevant materials in the mid-IR, adds up for bulk stretcher and compressor. Consequently, the accumulated TOD needs compensation by the pulse shaper. The TOD can reach very high values, since the femtosecond signal pulse duration is matched to the few-picosecond pump pulse duration. As a consequence, large amounts of bulk material have to be implemented in the OPCPA system to manage the group delay dispersion (GDD).
Here we report on exploring the limits of dispersion management with bulk materials in high-energy mid-IR OPCPAs and more importantly, of the feasible amount of TOD compensated by available shapers. For this study a few-cycle MWIR OPCPA is used which delivers pulse energies of 2 mJ and 3.4 mJ at 3.4 µm and at 4.9 µm respectively [22]. Shaping and compression of these pulses using different approaches is presented with the aim of generating high energy few-cycle pulses at 12 µm via DFG. Because signal and idler pulses are used for DFG, optimal phase control of both pulses is challenging since there is only one shaper in the MWIR OPCPA.
2. Pulse shaping issues in mid-IR OPCPAs – general remarks
The most convenient approach to imprint an almost arbitrary phase and amplitude on a mid-IR ultrashort pulse is the employment of a spatial-light modulator (SLM) [31] or an acousto-optic programmable dispersive filter (AOPDF) [32]. However, the usage of an SLM or an AOPDF is limited to pulses with low energy (few-µJ), and is therefore only suitable for seed pulses in high-energy OPCPA systems. Furthermore, the implementable dispersion (GDD, TOD) values decrease with increasing wavelength for both shaper types. Unfortunately, reflection type MEMS with a metallic mirror array, potentially providing the required stroke for shaping at longer wavelengths in the mid-IR, are not available.
It should be mentioned that for wavelengths longer than 3 µm GRISMs, reflection-type as well as transmission-type diffraction gratings with high efficiency (> 90%), are not available. Of course, ruled Au gratings with a typical diffraction efficiency of 80% remain, which lead to losses of >40% for a compressor arrangement, are thus not desirable. Chirped Mirrors (CM) are now available in the 3 µm wavelength range with a maximum GDD of ∼500 fs2 per bounce. However, for stretching the fs signal pulses to the required few-ps pulse duration in OPCPAs, the necessary GDD of >50,000 fs2 requires approx. 100 CMs. Thus, they are only useful for fine tuning of the dispersion [33].
In mid-IR OPCPAs, the use of bulk materials for stretching and compression seems to be straight forward because common transparent crystals exhibit negative (Al2O3 (transparent only up to 4.5 µm), CaF2, MgF2, BaF2) or positive (Si, Ge) group velocity dispersion (GVD). The stretching and compression ratios can be adjusted by the crystal thickness. However, the nonlinearity of the respective material has to be considered in the context of high-energy pulses compression. Si and Ge exhibit a large nonlinear refractive index n2 which can lead to severe pulse distortion.
In the parametric amplification process, the sign of the GDD is inverted between signal and idler pulses. Hence, if aiming for the idler pulse, the same material can in principle be used for stretching the signal and compressing the idler after amplification. Accordingly, the pulse duration of the signal or the idler can be managed for each subsequent amplification stage in the OPCPA individually.
The TOD, however, does not revert in the parametric process and is positive in the mid-IR for all the materials listed above. Consequently, the TOD sums up for bulk stretchers and compressors and increases significantly for most materials with wavelength. Exceptions are Ge and Si, where the TOD is relatively constant with a moderate value over the entire mid-IR spectral range.
This leads to an essential limitation of the bulk/shaper combination for the dispersion management in the mid-IR: The maximum TOD being compensated with the shaper limits the amount of the material and thus, the maximum temporal stretching of the pulses. Typical TOD values manageable with both shaper types are in the range of ∼1 x 106 fs3 [34,35].
3. Experimental setup
The experimental scheme is presented in Fig. 1(a). The few-cycle mid-wave IR-OPCPA system operates at 1 kHz repetition rate and is described in detail in [22]. The 40 MHz front-end, based on a Er:fiber laser (Toptica), provides femtosecond pulses at 1.0 µm, 1.5 µm and 2.0 µm wavelength. The latter are the seed pulses for the pump whereas the 1.0 µm and 1.5 µm pulses generate the signal pulses at 3.5 µm by means of DFG in a periodically poled lithium niobate crystal (MgO:PPLN). The DFG exhibits a duration of 30 fs with 0.5 pJ energy. The parametric amplifier consists of four stages which are pumped at 2.05 µm by a Ho:YLF chirped pulse amplifier (CPA) system [19]. AR-coated ZnGeP2 (ZGP) crystals (BAE Systems) are used as nonlinear crystals in all four OPA stages. In [22], the last booster stage was still pumped with the residual pump after the third stage. In order to increase the stability of the system, it is now pumped directly like all other stages (Fig. 1(a)). Except for the fourth OPA stage, all previous stages are seeded with the signal. Only the third stage has a collinear design to seed the last, non-collinear stage with the idler at 5 µm (Fig. 1(a)). Thus, the signal pulses after the fourth stage exhibit an angular dispersion and consequently, the signal pulses at 3.5 µm after the third stage are used for the aimed LWIR DFG experiments.
Fig. 1. (a) Setup of the MWIR OPCPA and the LWIR difference frequency generation (DFG). It comprises the 3-color front-end, the 2.05 µm Ho:YLF chirped pulse amplifier (CPA) as pump, the four optical parametric amplifier (OPA) stages based on ZGP crystals. SLM, spatial light modulator; DM, dichroic mirrors; LF, long-pass filter; S, bulk stretcher (CaF2 or Al2O3). (b-d) Different pulse shaping configurations used in this work: spectral shaper + OPCPA (last stage) + Compressor; color code: red – pump, green – signal, violet – idler, black – LWIR DFG.
Prior to the parametric amplification, adaptive phase control of the signal pulses is conducted by a spatial light modulator (SLM). The shaper design is a 4-f setup with a 600 lines/mm grating as the dispersive element (Fig. 1(b), left). The SLM is a 512×512-pixel liquid crystal (LC) array with a pixel pitch of 25 µm applicable in the spectral range from 2.8 µm to 4.0 µm (Meadowlark Optics).
For the characterization of the compressed pulses emitted by the MWIR OPCPA, a home-built second harmonic frequency-resolved optical gating (SH-FROG) device is used. AgGaS2 and GaSe are used as nonlinear crystals for generating the SH. A near-infrared spectrometer (NIRQuest, Ocean Insight) and a scanning monochromator (iHR320, Horiba) serve as spectrometers for the signal and idler pulse characterization, respectively.
4. Results and discussion
4.1 Idler pulses at 4.9 µm – SLM shaping and CaF2 compression
As a starting point for our considerations, we first briefly review the output performance of our system described in [22].
If aiming at the 4.9 µm idler, the pulses should be positively chirped to be able to use CaF2 bulk material for re-compression. This condition determines the dispersion design in the MWIR OPCPA. Due to the inherent sign inversion of the GDD between signal and idler in OPCPAs, the signal pulses at 3.4 µm have to be negatively chirped. In our system, CaF2 and Al2O3 rods are used to stretch the pulses to 2.5 ps duration. Compression of the positively chirped idler pulses is performed in CaF2 rods placed under Brewster angle. The accumulated TOD of ∼0.9×106 fs3 is pre-compensated by indirect phase shaping of the signal pulses with the SLM in connection with fine tuning of the residual GDD. The TOD introduced into our OPCPA system thus borders on the limit of what the SLM can compensate [34]. The detailed dispersion values for this configuration are summarized in Table 1.
Table 1. Dispersion values for the signal at 3.4 µm and idler pulses at 4.9 µm in the MWIR OPCPA with SLM and bulk compression (OPCPA dispersion is given by the stretcher and the OPA chain, τp denotes the resulting recompressed pulse duration)
Figure 2 shows the SH-FROG characterization of the idler pulses with an SLM as the shaper and re-compression in four 25 mm-thick CaF2 rods placed at Brewster’s angle. The idler spectrum covers a range from 4.2 to 5.4 µm (at 1/e2) supporting a Fourier transform limited (FTL) pulse duration of 70 fs (Fig. 2(c)). The pulse retrieval delivers a temporal shape with a duration 85 fs confirming the third harmonic-FROG measurement in [22]. Effectively pumped with 33 mJ, an idler pulse energy of 3.4 mJ is generated, which is still the highest value of among fs mid-IR sources beyond 4 µm [6]. These pulses are routinely used for the generation of hard X-ray pulses with exceptional high flux at 1 kHz [36].
Fig. 2. SH-FROG characterization of the idler pulse at 4.9 µm with CaF2 bulk compression and SLM for a pulse energy of 3.4 mJ. (a), (b) SH-FROG trace measured and retrieved, FROG error: 0.5%; (c) optical spectrum, measured (grey), retrieved (violet) and phase (red); (d) retrieved temporal pulse shape for the SLM in operation and switched off.
To demonstrate the impact of the SLM on pulse the performance, it was turned off. The resulting uncompensated TOD is clearly reflected in the retrieved pulse by the now occurring structured temporal pulse shape and the significantly longer duration of 120 fs, see Fig. 2(d).
4.2 Signal pulses at 3.4 µm – SLM shaping and Si compression
For the aimed few-cycle pulse generation at 12 µm via DFG, the signal pulses at 3.4 µm have also to be compressed. As the signal is negatively chirped, it requires positive GDD and negative TOD to compensate the phase difference. At first we tested the appealing option to use silicon for bulk compression. Si bulk compression was reported for µJ energy pulses around 3 µm. 125-µJ pulses at 3.1 µm were compressed to ∼170 fs with Si and finally to 73 fs using chirped mirrors [33]. Si compression without chirped mirrors to 38 fs with a pulse energy of 150 µJ at 3.1 µm was achieved [37]. Compression of pulse energies >5 mJ at 3 µm using Si was proposed [38].
Besides the expected challenges with the high n2 of Si, the absorption in Si must be kept as low as possible when operating at high average power to prevent thermal problems. The amplified signal pulse energy after the third stage amounts to 2 mJ. This results in 2 W average power at 3.4 µm at a 1 kHz repetition rate. The absorption of Si depends on the doping level, e.g., phosphorus, and potential impurities. A measure of the doping level is the resistivity R (Ω x cm). The absorption (1/cm) in Si increases approximately by one order of magnitude if the resistivity is reduced by one order of magnitude [39]. Thus, undoped silicon with high resistivity and nearly no impurities is the most suitable for our application. For this purpose, a high purity 4“ diameter Si crystal with resistivity not less than 8,000 Ω x cm from the Leibniz-Institut für Kristallzüchtung, Berlin (IKZ) was applied. The crystal was grown by floating zone technique in (100) direction and has n-type conductivity. Si plates with (100) surface orientation were cut and polished using standard chemical-mechanical polishing.
The Si compressor consists of four plates with the required overall thickness of 70 mm. The plates are placed at Brewster’s angle. The successful re-compression of the signal pulses is presented in Fig. 3. The signal spectrum extends from 3.1 µm to 3.9 µm (Fig. 3(c)). This bandwidth supports a 40 fs FTL pulse duration. The dip in the spectrum at 3.4 µm originates from the absorption characteristics of the used liquid in the LCOS SLM array [40]. The SH-FROG characterization delivers a retrieved pulse duration for the signal as short as 45 fs. The corresponding settings of the SLM and the other dispersion parameters in this configuration are listed in Table. 1. However, a clean pulse compression in the Si plates is limited to maximum pulse energy of 40 µJ where beam distortions become a problem. The latter is illustrated by monitoring the profile of the propagating beam before and after the Si compressor (Fig. 4). The intensity distribution is shown at four distinct positions for a field-of-view of 10.6 mm x 10.6 mm in Fig. 4(a)-(d), indicating the effect of the Kerr-lens which leads to focusing of the beam. Due to the Brewster geometry, the beam cross sections in the Si plates are different in the two planes, so that the focal distances as well as the focal size differ in the x- and y-direction (Fig. 4(e)). From the fits in Fig. 4(e), a beam quality factor M2 of 1.2 was estimated for both planes showing no signs of deterioration during compression. The measured focal lengths of 0.8 m and 1.7 m for the x- and y-direction, respectively, are in good agreement with the calculated focal lengths using n2 = 4.54×10−14 cm2/W at 3.2 µm for Si [41] with a pulse intensity in x-plane of 6 GW/cm2. Consequently, Si compression of the 3.4 µm pulses was not pursued. It is clear that the Brewster geometry for Si is unsuitable because it leads to beam astigmatism already for few-µJ pulse energies and if placed perpendicular to the beam AR-coating is essential. More important are the resulting design requirements when using Si as bulk compressor. For our 2 mJ signal pulses a beam diameter larger than 40 mm is necessary to prevent strong focussing and an accumulation of the B-integral. This requires at least 4''-size Si wafers and correspondingly large optics for beam shaping which is outside of our table-top approach for the OPCPA system.
Fig. 3. SH-FROG characterization of the signal pulse at 3.4 µm with Si bulk compression and SLM for a pulse energy of 40 µJ. (a), (b) SH-FROG trace measured and retrieved, FROG error: 0.3%; (c) optical spectrum, measured (grey), retrieved (green) and phase (turquoise); (d) retrieved temporal pulse shape and phase (turquoise).
Fig. 4. Beam profile of the propagating signal pulse at 3.4 µm before and after the Si compressor. (a-d) Intensity distribution at four distinct positions z: field-of-view of 10.6 mm x 10.6 mm. (e) Measured beam waists in dependence on the propagation distance the x- and y-planes and the corresponding fits.
At this point in our experiments, the SLM was no longer working properly, i.e., it was damaged. The spectral phase of the pulses is no longer correctly shaped which is to some extend already visible in the pulse spectrum above 3.6 µm (Fig. 3(c)). Furthermore, it was not possible to achieve a residual phase close to zero for the compressed pulses, as can be seen from the dispersion values listed for this configuration in Table. 1. Since the resulting phase has a negative GDD and positive TOD, a contribution from SPM cannot be ruled out. We noticed that the maximum phase stroke of the SLM is reduced so that 2π phase shifts could not be realized anymore which finally leads to an obscure phase response behaviour. A solution had to be found for this problem as well. Active shapers were not on hand and the alternative had to provide a high negative TOD. With these constraints, we decided on a Ge-prism pair setup.
4.3 Idler pulses at 4.9 µm – Ge-prism pair shaping and CaF2 compression
For the targeted LWIR DFG experiments, a high conversion efficiency is aimed. For this purpose, the two driver pulses should have a short and almost the same duration. In our MWIR OPCPA, the compressed idler pulses are about twice as long as the signal pulses, see Table. 1. Therefore, the settings of the pulse shaper in the OPCPA for the LWIR DFG are chosen aiming for the shortest idler pulse. An adjustment of the signal pulse duration to that of the idler pulses should thus be possible.
The SLM was replaced by a pair of Ge Brewster prisms for shaping the signal pulses with 50 mm side lengths (Fig. 1(c),(d) left). With the knowledge of the GDD and TOD parameters for the shortest idler pulses (Tab. 1), the prism arrangement is configured to get as close as possible to these values. Due to this rebuilt and the required beam adaptation in the MWIR OPCPA, pulse energies of 0.7 mJ and 1.9 mJ are still available for signal and idler, respectively.
The SH-FROG characterization of the s-polarized idler pulses, shaped with the Ge-prism pair and CaF2 bulk compressor is shown in Fig. 5. The measured spectrum ranges from 4.6 to 5.2 µm which is smaller than the spectrum with the SLM shaper setup (explanation see section 4.4) and would allow a FTL pulse duration of 80 fs (Fig. 5(c)). Due to the modifications in the OPCPA design, more material is required for the idler compression. With a total thickness of 120 mm CaF2, shortest pulses with a duration of 120 fs are achieved, as shown by the SH-FROG retrieval in Fig. 5(d). The main pulse contains 95% of the energy and the spectral phase is practically flat despite the manipulations on the system. The dispersion values of this configuration are listed in Table. 2. With a residual TOD of +0.2×106 fs3, the dispersion management for the idler is not perfect which we attribute to the extremely difficult fine-tuning of the individual dispersion orders with the Ge prism pair.
Table 2. Dispersion values for the signal at 3.4 µm and idler pulses at 4.9 µm in the MWIR OPCPA with Ge prism pair as pulse shaper (OPCPA dispersion is given by the stretcher and the OPA chain, τp denotes the resulting recompressed pulse duration)
Fig. 5. SH-FROG characterization of the idler pulse at 4.9 µm with CaF2 bulk compression and Ge-prism pair as shaper for a pulse energy of 1.9 mJ. (a), (b) SH-FROG trace measured and retrieved, FROG error: 0.5%; (c) optical spectrum, measured (grey), retrieved (violet) and phase (red); (d) retrieved temporal pulse shape (violet) and phase (red).
4.4 Signal pulses at 3.4 µm – Ge-prism pair shaping and Martinez-type compressor
Next step is the compression of the signal pulses at 3.4 µm for the LWIR DFG. Our attempt of using Si bulk for compression cannot be applied here due to the high pulse energy, see Fig. 4. Therefore, an alternative to Si material is required to compress the p-polarized signal pulses.
Since the values of the Ge-prism shaper are predefined to ensure the shortest idler pulses, the alternative approach should allow the variation of GDD and TOD values of the signal pulse compressor. The goal is to get as close as possible to the idler pulse duration of 120 fs. Thus, a Martinez-type compressor [42] in a 4-f arrangement containing two CaF2 cylindrical lenses and two sapphire prisms as dispersive elements was used (Fig. 1(d), right).
Figure 6 shows the results of the signal compression using the Ge-prism shaper and the Martinez-type compressor (Fig. 1(d)). The spectrum is centered at 3.45 µm (Fig. 6(c)) with a bandwidth significantly narrower than with the SLM shaper (Fig. 3(c)). The bandwidth of the signal spectrum is limited to 300 nm (zero level) by the 50 mm side length of the Ge prisms (Fig. 6(c)). However, the spectrum still supports a FTL pulse duration of 100 fs. In contrast to the signal spectrum in Fig. 3(c), it no longer shows a distinct structuring, which is due to the lack of the SLM.
Fig. 6. SH-FROG characterization of the signal pulse at 3.4 µm with a Ge-prism pair as shaper and a Martinez-type compressor for a pulse energy of 0.7 mJ. (a), (b) SH-FROG trace measured and retrieved, FROG error: 0.3%; (c) optical spectrum, measured (grey), retrieved (green) and phase (turquoise); (d) retrieved temporal pulse shape (green) and phase (turquoise).
The SH-FROG characterization performed again, retrieves a compressed signal pulse duration of 125 fs (Fig. 6(d)). The goal is almost achieved being close to the idler pulse duration of 120 fs, but with a non-perfect pulse quality. The latter manifests in Tab. 2 with a residual TOD of -0.8×106 fs and reflects in the temporal pulse shape in Fig. 6(d). The signal pulse has a prominent leading satellite at -200 fs containing 15% of the total energy (Fig. 6(d)). This part of the pulse therefore cannot contribute to the LWIR DFG process.
4.5 LWIR difference frequency generation at 12 µm – Ge bulk compression
The results on LWIR DFG generation have recently been published as part of the work presented here [25]. The experiments taken from [25] complete the presented results in this paper.
Due to losses in the compressors and in the beam path, the pulse energies applied to the DFG crystal were 0.35 mJ and 1.0 mJ at 3.4 µm and 4.9 µm, respectively. Considering the pulse shape of both main pulses, the total energy available for the generation of the LWIR DFG is 1.25 mJ. A 1-mm thick silver gallium selenide (AGSe) crystal is used for the LWIR DFG. The AGSe crystal is cut at an angle of 53.3° and the phase-matching angle is 55.8° (Type-II). The beam diameter on the AGSe crystal is 4.3 mm × 3.6 mm. The resulting peak intensity is slightly below 200 GW/cm2 allowing safe operation without crystal damage.
The beams are combined with a dichroic mirror (Fig. 1(a)) and the LWIR DFG pulses are characterized behind a long-pass filter (Fig. 1(a)). A maximum pulse energy behind the long pass filter of 12.2 µJ is achieved, corresponding to an external conversion efficiency of >1%. The LWIR DFG spectrum is centered at 12.2 µm with a bandwidth of 1.74 µm (FWHM) (Fig. 7, inset) yielding a FTL pulse duration of slightly below 100 fs. Table 2 shows that both driver pulses are slightly chirped, however, the dispersion of the LWIR DFG pulses is dominated by the dichroic mirror in front of the nonlinear crystals (Fig. 1(a)). This mirror imposes a GDD of -6,000 fs2 on the transmitted 3.4-µm pulses and hence requires post compression. This is performed using three AR-coated Ge plates with a total thickness of 24 mm accompanied by some losses. The LWIR DFG pulse characterization is performed with an autocorrelator (pulseCheck, APE GmbH), operating in the TPA mode for wavelengths beyond 10 µm, and showed a pulse durations of 143 fs, assuming a sech2-shape (Fig. 7). We attribute the deviation from the FTL duration to the uncompensated TOD imprinted on the 3.4-µm pulses (see Tab. 2). Hence, the generated DFG pulses have a duration of less than four optical cycles and a pulse energy as high as 10 µJ. The intensity in the Ge compressor amounts to 1.3 GW/cm2. At this intensity no beam distortions are observable. The beam profile is nearly Gaussian, measured using a Pyrocam (IIIHR, Ophir), and the power stability is remarkably at 1% rms, measured at 10 mW average power [25].
Fig. 7. Pulse characterization of the generated DFG at 12 µm in AGSe. Measured autocorrelation trace and spectrum (inset).
4.6 Idler pulses tunable from 5.4 to 6.8 µm – AOPDF shaping and CaF2 compression
In order to classify the presented SLM pulse shaping performance, we compare it to the AOPDF pulse shaping, which we performed in a tunable high-energy midwave OPCPA [23]. The latter is also pumped at 2.05 µm by a Ho:YLF regenerative amplifier and the parametric amplifier architecture is structurally similar to the setup in Fig. 1(a), but contains a different front end (fs Cr:ZnS laser). After amplification in two stages, also based on ZGP crystals, the idler pulses reached more than 400 µJ of energy at 1 kHz repetition rate. They were wavelength tunable between 5.4 and 6.8 µm with a pulse duration of less than 100 fs. Prior to the amplifier chain, the signal pulses were shaped by an AOPDF (Dazzler, Fastlite). For compression CaF2 bulk material was introduced using an adjustable prism pair. To enable the shortest pulse duration in the case of 5.4 µm center wavelength, 46 mm CaF2 was needed and AOPDF settings of +3,000 fs2 GDD and +600 x 103 fs3 TOD. The maximum TOD value that could be implemented in the AOPDF was ∼1 x 106 fs3. In our system, we achieved a maximum shaper throughput of 20% for the corresponding signal pulses at 3.2 µm [23].
5. Conclusion
Dispersion management in a multi-mJ few-cycle MWIR OPCPA using combinations of bulk materials and spectral pulse shapers was presented. The generation of sub-five optical cycle pulses for the signal and the idler was demonstrated at 3.4 µm and 4.9 µm, respectively, using SLM pulse shaping optimized for each center wavelength. The need to replace the SLM as an active phase shaping device led us to introduce a Ge-prism pair as an alternative. The generated signal and idler pulses were somewhat longer compared to the SLM configuration due to the fine-tuning of the spectral phase that was no longer possible with the Ge-prism pair and its reduced transmitted bandwidth. However, the dispersion could be managed in such a way that signal and idler pulses had almost the same duration of ∼120 fs simultaneously.
These pulses were used for difference frequency generation in AGSe at a 1 kHz repetition rate. Pulses at 12 µm wavelength with 10 µJ energy and 143 fs duration are achieved corresponding to sub-four optical cycles. Simulations predict a sub-100 fs duration of the LWIR DFG pulses with multi-10 µJ energy in our setup if pulse shaping of signal and idler is performed using the SLM instead of the Ge-prism pair, i.e., pulse durations of less than 100 fs can be provided from the MWIR OPCPA for the signal and idler pulses.
The SLM shaping performance in the midwave-IR OPCPA was compared to that of an AOPDF in a similar system, a tunable high-energy midwave ZGP OPCPA described in section 4.6. Using the AOPDF, maximum TOD values for the signal pulses at 3.2 µm of ∼1 x 106 could be set. A certain amount of GDD increases the diffraction efficiency considerably and a pure TOD compensation is not advisable. Consequently, a certain GDD and TOD combination provides the largest seed energy for the OPCPA. An increased GDD of the AOPDF reduces the necessary material in the setup and thus also the TOD. In contrast, when using the SLM, there are no dependencies between the settings of the GDD and the TOD, they can be managed independently. As with the AOPDF, the maximum value of the TOD to be set is approx. 1 x 106 fs for the SLM at the same signal wavelength. An advantage of the SLM shaper design is that, despite its implementation in a 4-f-setup, a higher throughput can be achieved compared to the AOPDF. In our midwave-IR OPCPA setups the SLM shaper provides a throughput of 40% for the signal pulse whereas this value is only 20% for the AOPDF.
Funding
European Regional Development Fund (CZ.02.1.01/0.0/0.0/15_006/0000674); Horizon 2020 Framework Programme (739573, 871124); Deutsche Forschungsgemeinschaft (GR2115/6-1).
Acknowledgments
Expert technical support of Peggy Hofmann (CrysTec) is gratefully acknowledged.
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. B. Wolter, M. G. Pullen, M. Baudisch, M. Sclafani, M. Hemmer, A. Senftleben, C. D. Schröter, J. Ullrich, R. Moshammer, and J. Biegert, “Strong-Field Physics with Mid-IR Fields,” Phys. Rev. X 5(2), 021034 (2015). [CrossRef]
2. D. Popmintchev, B. R. Galloway, M.-C. Chen, F. Dollar, C. A. Mancuso, A. Hankla, L. Miaja-Avila, G. O’Neil, J. M. Shaw, G. Fan, S. Ališauskas, G. Andriukaitis, T. Balčiunas, O. D. Mücke, A. Pugzlys, A. Baltuška, H. C. Kapteyn, T. Popmintchev, and M. M. Murnane, “Near- and Extended-Edge X-Ray-Absorption Fine-Structure Spectroscopy Using Ultrafast Coherent High-Order Harmonic Supercontinua,” Phys. Rev. Lett. 120(9), 093002 (2018). [CrossRef]
3. A. Ashihara, N. Huse, A. Espagne, E. T. J. Nibbering, and T. Elsaesser, “Ultrafast structural dynamics of water induced by dissipation of vibrational energy,” J. Phys. Chem. A 111(5), 743–746 (2007). [CrossRef]
4. H. Fattahi, H. G. Barros, M. Gorjan, et al., “Third-generation femtosecond technology,” Optica 1(1), 45 (2014). [CrossRef]
5. H. Pires, M. Baudisch, D. Sanchez, M. Hemmer, and J. Biegert, “Ultrashort pulse generation in the mid-IR,” Prog. Quantum Electron. 43, 1–30 (2015). [CrossRef]
6. A. Dubietis and A. Matijošius, “Table-top optical parametric chirped pulse amplifiers: past and present,” Opto-Electron. Adv. 6(3), 220046 (2023). [CrossRef]
7. M. Ferray, A. L’Huillier, X. F. Li, L. A. Lompre, G. Mainfray, and C. Manus, “Multiple-harmonic conversion of 1064 nm radiation in rare gases,” J. Phys. B: At. Mol. Opt. Phys. 21(3), L31–L35 (1988). [CrossRef]
8. G. Ciriolo, M. Negro, M. Devetta, E. Cinquanta, D. Faccialà, A. Pusala, S. De Silvestri, S. Stagira, and C. Vozzi, “Optical parametric amplification techniques for the generation of high-energy few-optical-cycles IR pulses for strong field applications,” Appl. Sci. 7(3), 265 (2017). [CrossRef]
9. P. B. Corkum, “Plasma perspective on strong-field multiphoton ionization,” Phys. Rev. Lett. 71(13), 1994–1997 (1993). [CrossRef]
10. J. Weisshaupt, V. Juvé, M. Holtz, S. A. Ku, M. Woerner, T. Elsaesser, S. Ališauskas, A. Pugžlys, and A. Baltuška, “High-brightness table-top hard x-ray source driven by sub-100 femtosecond mid-infrared pulses,” Nat. Photonics 8(12), 927–930 (2014). [CrossRef]
11. A. M. Weiner, “Ultrafast optical pulse shaping: A tutorial review,” Opt. Commun. 284(15), 3669–3692 (2011). [CrossRef]
12. S. Demmler, J. Rothhardt, A. M. Heidt, A. Hartung, E. G. Rohwer, H. Bartelt, J. Limpert, and A. Tuennermann, “Generation of high quality, 1.3 cycle pulses by active phase control of an octave spanning supercontinuum,” Opt. Express 19(21), 20151–20158 (2011). [CrossRef]
13. B. Xu, Y. Coello, V. V. Lozovoy, D. A. Harris, and M. Dantus, “Pulse shaping of octave spanning femtosecond laser pulses,” Opt. Express 14(22), 10939–10944 (2006). [CrossRef]
14. K. Yamane, Z. Zhang, K. Oka, R. Morita, M. Yamashita, and A. Suguro, “Optical pulse compression to 3.4 fs in the monocycle region by feedback phase compensation,” Opt. Lett. 28(22), 2258–2260 (2003). [CrossRef]
15. B. Schenkel, J. Biegert, U. Keller, C. Vozzi, M. Nisoli, G. Sansone, S. Stagira, S. De Silvestri, and O. Svelto, “Generation of 3.8-fs pulses from adaptive compression of a cascaded hollow fiber supercontinuum,” Opt. Lett. 28(20), 1987–1989 (2003). [CrossRef]
16. A. Prakelt, M. Wollenhaupt, A. Assion, C. Horn, C. Sarpe-Tudoran, M. Winter, and T. Baumert, “Compact, robust, and flexible setup for femtosecond pulse shaping,” Rev. Sci. Instrum. 74(11), 4950–4953 (2003). [CrossRef]
17. V. Petrov, “Frequency down-conversion of solid-state laser sources to the mid-infrared spectral range using non-oxide nonlinear crystals,” Prog. Quantum Electron. 42, 1–106 (2015). [CrossRef]
18. P. G. Schunemann, K. T. Zawilski, L. A. Pomeranz, D. J. Creeden, and P. A. Budni, “Advances in nonlinear optical crystals for mid-infrared coherent sources,” J. Opt. Soc. Am. B 33(11), D36–D43 (2016). [CrossRef]
19. L. von Grafenstein, M. Bock, D. Ueberschaer, A. Koc, U. Griebner, and T. Elsaesser, “2.05 µm chirped pulse amplification system at a 1 kHz repetition rate - 2.4 ps pulses with 17 GW peak power,” Opt. Lett. 45(14), 3836–3839 (2020). [CrossRef]
20. U. Elu, T. Steinle, D. Sánchez, L. Maidment, K. Zawilski, P. Schunemann, U. D. Zeitner, C. Simon-Boisson, and J. Biegert, “Table-top high-energy 7 µm OPCPA and 260 mJ Ho:YLF pump laser,” Opt. Lett. 44(13), 3194–3197 (2019). [CrossRef]
21. L. von Grafenstein, M. Bock, D. Ueberschaer, K. Zawilski, P. Schunemann, U. Griebner, and T. Elsaesser, “5 µm, few-cycle pulses with multi-GW peak power at a 1 kHz repetition rate,” Opt. Lett. 42(19), 3796–3799 (2017). [CrossRef]
22. L. von Grafenstein, M. Bock, D. Ueberschaer, E. Escoto, A. Koç, K. Zawilski, P. Schunemann, U. Griebner, and T. Elsaesser, “Multi-millijoule, few-cycle 5µm OPCPA at 1 kHz repetition rate,” Opt. Lett. 45(21), 5998 (2020). [CrossRef]
23. P. Fuertjes, L. von Grafenstein, C. Mei, M. Bock, U. Griebner, and T. Elsaesser, “Cr:ZnS-based soliton self-frequency shifted signal generation for a tunable sub-100 fs MWIR OPCPA,” Opt. Express 30(4), 5142 (2022). [CrossRef]
24. F. Rotermund, V. Petrov, and F. Noack, “Difference-frequency generation of intense femtosecond pulses in the mid-IR (4-12 µm) using HgGa2S4 and AgGaS2,” Opt. Commun. 185(1-3), 177–183 (2000). [CrossRef]
25. M. Duda, L. von Grafenstein, M. Bock, D. Ueberschaer, P. Fuertjes, L. Roškot, M. Smrž, O. Novák, and U. Griebner, “10-µJ few-cycle 12-µm source based on difference-frequency generation driven by a 1-kHz mid-wave infrared OPCPA,” Opt. Lett. 47(11), 2891 (2022). [CrossRef]
26. F. Junginger, A. Sell, O. Schubert, B. Mayer, D. Brida, M. Marangoni, G. Cerullo, A. Leitenstorfer, and R. Huber, “Single-cycle multiterahertz transients with peak fields above 10 MV/cm,” Opt. Lett. 35(15), 2645–2647 (2010). [CrossRef]
27. M. Babzien, I. V. Pogorelsky, and M. Polanskiy, “Solid-state seeding of a high power picosecond carbon dioxide laser,” AIP Conf. Proc. 1777, 110001 (2016). [CrossRef]
28. K. Liu, H. Liang, L. Wang, S. Qu, T. Lang, H. Li, Q. J. Wang, and Y. Zhang, “Multimicrojoule GaSe-based midinfrared optical parametric amplifier with an ultrabroad idler spectrum covering 4.2–16 µm,” Opt. Lett. 44(4), 1003–1006 (2019). [CrossRef]
29. R. Budriunas, K. Jurkus, M. Vengris, and A. Varanavicius, “Long seed, short pump: converting Yb-doped laser radiation to multi-µJ few-cycle pulses tunable through 2.5–15 µm,” Opt. Express 30(8), 13009 (2022). [CrossRef]
30. P. Fuertjes, M. Bock, L. von Grafenstein, D. Ueberschaer, U. Griebner, and T. Elsaesser, “Few-cycle 65 µJ pulses at 11.4 µm for ultrafast longwave-infrared spectroscopy,” Optica 9(11), 1303–1306 (2022). [CrossRef]
31. S. Serati, T. K. Ewing, and J. Stockley, “New developments in high-resolution liquid-crystal spatial light modulators for wavefront control,” Proc. SPIE 4825, 46–55 (2002). [CrossRef]
32. R. Maksimenka, P. Nuernberger, K. F. Lee, A. Bonvalet, J. Milkiewicz, C. Barta, M. Klima, T. Oksenhendler, P. Tournois, D. Kaplan, and M. Joffre, “Direct mid-infrared femtosecond pulse shaping with a calomel acousto-optic programmable dispersive filter,” Opt. Lett. 35(21), 3565 (2010). [CrossRef]
33. M. Mero, Z. Heiner, V. Petrov, H. Rottke, F. Branchi, G. M. Thomas, and M. J. J. Vrakking, “43 W, 1.55 µm and 12.5 W, 3.1 µm dual-beam, sub-10 cycle, 100 kHz optical parametric chirped pulse amplifier,” Opt. Lett. 43(21), 5246–5249 (2018). [CrossRef]
34. M. Bock, L. von Grafenstein, U. Griebner, and T. Elsaesser, “Generation of millijoule few-cycle pulses at 5 µm by indirect spectral shaping of the idler in an optical parametric chirped pulse amplifier,” J. Opt. Soc. Am. B 35(12), C18 (2018). [CrossRef]
35. R. Maksimenka and P. Tournois, “Mid-infrared high-frequency high-resolution reflective acousto-optic filters in mercury halides,” Opt. Commun. 285(5), 715–719 (2012). [CrossRef]
36. A. Koç, C. Hauf, M. Woerner, L. von Grafensteın, D. Ueberschaer, M. Bock, U. Griebner, and T. Elsaesser, “Compact high-flux hard X-ray source driven by femtosecond mid-infrared pulses at a 1 kHz repetition rate,” Opt. Lett. 46(2), 210–213 (2021). [CrossRef]
37. N. Thiré, R. Maksimenka, B. Kiss, C. Ferchaud, G. Gitzinger, T. Pinoteau, H. Jousselin, S. Jarosch, P. Bizouard, V. Di Pietro, E. Cormier, K. Osvay, and N. Forget, “Highly stable, 15 W, few-cycle, 65 mrad CEP-noise mid-IR OPCPA for statistical physics,” Opt. Express 26(21), 26907–26915 (2018). [CrossRef]
38. Y. Yin, J. Li, X. Ren, Y. Wang, A. Chew, and Z. Chang, “High-energy two-cycle pulses at 3.2 µm by a broadband-pumped dual-chirped optical parametric amplification,” Opt. Express 24(22), 24989–24998 (2016). [CrossRef]
39. J. Degallaix, R. Flaminio, D. Forest, M. Granata, C. Michel, L. Pinard, T. Bertrand, and G. Cagnoli, “Bulk optical absorption of high resistivity silicon at 1550 nm,” Opt. Lett. 38(12), 2047 (2013). [CrossRef]
40. T. K. Ewing and W. R. Folks, “Liquid crystal on silicon infrared scene projectors,” Proc. SPIE 5785, 36–45 (2005). [CrossRef]
41. G. Jansonas, R. Budriūnas, M. Vengris, and A. Varanavičius, “Interferometric measurements of nonlinear refractive index in the infrared spectral range,” Opt. Express 30(17), 30507 (2022). [CrossRef]
42. O. E. Martinez, “3000 times grating compressor with positive group velocity dispersion: Application to fiber compensation in 1.3-1.6 µm region,” IEEE J. Quantum Electron. 23(1), 59–64 (1987). [CrossRef]
