Open Access
Add to BibSonomy
Abstract
A compact and robust all-fiber difference frequency generation-based source of broadband mid-infrared radiation is presented. The source emits tunable radiation in the range between 6.5 µm and 9 µm with an average output power up to 5 mW at 125 MHz repetition frequency. The all-in-fiber construction of the source along with active stabilization techniques results in long-term repetition rate stability of 3 Hz per 10 h and a standard deviation of the output power better than 0.8% per 1 h. The applicability of the presented source to laser spectroscopy is demonstrated by measuring the absorption spectrum of nitrous oxide (N2O) around 7.8 µm. The robustness and good long- and short-term stability of the source opens up for applications outside the laboratory.
© 2019 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Recent decade brought tremendous progress in laser-based gas detection techniques. One of the main driving factors behind the evolution of this particular sensing method was the relentless devotion of the scientific community to develop state-of-the-art sources capable of targeting strong absorption lines of gases localized in the mid-infrared (mid-IR) wavelength region. Traditional absorption detection techniques based on narrowband, semiconductor mid-IR lasers offer non-complex implementation at a cost of narrow tuning range, usually limited to several wavenumbers [1]. This drawback is especially evident in applications requiring simultaneous monitoring of several gas species, e.g. environmental monitoring [2]. In such systems individual expensive quantum cascade laser (QCL) or interband cascade laser (ICL) sources are typically incorporated into a single platform [3]. Other mid-IR sources addressing this issue have been proposed recently, including QCL combs [4,5], supercontinuum [6,7], Kerr-combs [8,9] or nonlinear frequency conversion sources [10–12]. Although QCL combs combine compactness and efficiency, their optical bandwidth is too low for some applications and no straightforward control of the repetition frequency (frep) or the carrier-envelope offset frequency (fCEO) had been proposed. One method of addressing these limitations is based on using nonlinear frequency conversion sources, which offer broadband, tunable coherent radiation in the mid-IR. Optical parametric oscillators (OPO) or difference frequency generation (DFG) are commonly used in combination with reliable and cost effective near infrared (near-IR) laser sources [13–15]. Recent development of reliable mid-IR comb sources, accompanied by clever methods of broadband detection and signal processing allowed for designing gas sensors capable of simultaneous monitoring of several gas analytes with comb-tooth resolution [16]. Broadband detection schemes compatible with mid-IR comb sources include e.g. Fourier-transform spectroscopy [17], virtually imaged phased array (VIPA) imaging [18], Vernier cavity filtering [19], or, very recently, dual-comb spectroscopy [16,20].
OPOs are the most common choice for comb spectroscopy in the mid-IR since they provide high output power with broad spectral coverage [21,22]. However, in contrast to OPOs, the DFG approach to generate mid-IR radiation is much more flexible – it does not require a cavity, therefore it can be implemented almost fully using optical fibers (which ensures alignment-free operation and reduced complexity). Moreover, the fully passive cancellation of fCEO makes controlling the comb frequencies straightforward. Various types of nonlinear crystals enable extremely broad spectral coverage of mid-IR DFG sources, spanning up to 5 µm with periodically poled lithium niobate (PPLN) [23,24], 12.7 µm with GaAs [25], 17 µm with AgGaSe2 [26], or 18 µm with GaSe using the intra-pulse DFG approach [27]. However, the previously reported DFG sources have limited applicability in spectroscopic systems due to the use of free-space optics (which requires periodic alignment and maintenance) [26,28,29], lack of repetition frequency tuning [23,30], and lack of active pulse overlap stabilization [25,28,30]. Especially the latter is critical for spectroscopic applications, since any mismatch in the temporal overlap between the two interacting pulses causes output power drifts and significant increase of intensity noise in the output beam, as discovered experimentally in [31].
Here we present a fully-fiberized, DFG-based mid-IR optical frequency comb source with 125 MHz repetition rate, capable of covering the 6.5 µm – 9 µm wavelength region, which addresses the main shortcomings of the previously reported DFGs. The source is designed to have a widely tunable and stabilized repetition frequency and actively stabilized overlap between the femtosecond pulses taking part in the nonlinear process. The delay stabilization ensures not only a stable output pulse amplitude, but also minimizes the intensity noise present in the idler beam. The applicability of the presented source was verified by measuring absorption spectra of nitrous oxide (N2O), a harmful greenhouse gas, at 7.5 – 8.1 µm. Robust, entirely fiberized and simple design paves the way to outside-lab applications of mid-IR DFG frequency combs, e.g. in environmental monitoring and broadband multi-species detection.
2. The DFG source layout and output
The schematic of the source is depicted in Fig. 1. It is composed of four main parts: 1) a mode-locked (ML) Er:fiber seed laser, 2) a 1.56 µm amplification stage, 3) a stage for achieving a soliton shift and subsequent amplification of the ∼2 µm red-shifted part of the spectrum, and 4) difference frequency generation. The ML seed laser was constructed as a simple and robust ring cavity as depicted in detail on the schematic in Fig. 2(a). A single hybrid fiber component serving as a wavelength division multiplexer, isolator and a 10% output coupler was used to minimize the complexity of the resonator. A 300 mW single-mode pump diode was used to optically pump a 43 cm-long erbium doped active fiber (Liekki Er80-4/125-PM) providing the necessary gain. The cavity was 1.63 m long with a net anomalous dispersion that allowed to mode-lock the laser in the soliton regime and generate 320 fs pulses [see Fig. 2(b)] at a 125.07 MHz repetition frequency [see Fig. 2(c)]. The mode-locking was provided by a graphene saturable absorber (GSA) consisting of 37 layers of graphene sandwiched between two fiber connectors [32]. The average power achieved at the 10% tap output was 5 mW.
Fig. 1. Schematic of the fiber-based Mid-IR comb source. PD – piezo driver, PID – PID controller, Mix – RF mixer, LO – local oscillator, FC – fiber coupler, DCF – dispersion compensating fiber, EDFA/TDFA – erbium/thulium doped fiber amplifier, PMF – polarization maintaining single-mode fiber, ODL – fiberized optical delay line, PZT – piezo-ceramic fiber stretcher, LA – logarithmic amplifier, BPF – electronic band-pass filter, DET – mid-IR detector, PCF – photonic crystal fiber, WDM – wavelength division multiplexer, COLL – collimator, FL – focusing lens, OPGaP - 3-mm long orientation patterned gallium phosphide crystal, BS – beam splitter. Electrical connections are shown in gray. Signal processing blocks responsible for stabilization are highlighted in dashed boxes.
Fig. 2. (a) The mode-locked seed laser setup, (b) the optical spectrum in a linear scale (inset shows the autocorrelation trace of the generated pulse), (c) the RF spectrum of the fundamental frep beatnote (inset shows the RF spectrum from DC to 6 GHz).
The pulses out-coupled from the ML seed laser are split by a 50/50 coupler and directed to separate sections of the DFG source. The upper stage on Fig. 1 (with blue shading) is a standard chirped pulse amplifier in which chirping, amplification and compression occurs only in optical fibers [33]. The pulse is first stretched in an in-house fabricated polarization maintaining (PM) dispersion compensating fiber (DCF) with negative dispersion at 1560 nm (−45 ps/nm/km). The chirped pulses are subsequently amplified in a standard erbium doped fiber amplifier (EDFA) and recompressed in a standard PM1550 fiber. At maximum pump power delivered to the amplification stage (1.5 W) the 1.56 µm pulses registered at the output of the common WDM coupler had a duration of 65 fs and average output power of 200 mW. An additional fiber-pigtailed optical delay line (ODL) with 80 mm travel and a PZT-based fiber stretcher with approx. 80 µm displacement were added to this section to enable both coarse optical path-length tuning and fast stabilization of the pulse overlap in the DFG crystal. The second stage of this setup (red-shaded in Fig. 1) was responsible for shifting the 1.56 µm pulses into the 2 µm wavelength region. The pulses originating from the ML seed laser were first amplified and recompressed in a dispersion tailored EDFA, reaching an average output power of approx. 300 mW and 50 fs pulse duration (see Fig. 3(a) for detailed output power vs. pump power characteristic of the amplifier). The amplified pulses were then coupled to a silica photonic crystal fiber (PCF) in which a Raman soliton self-frequency shift effect [34] was obtained. The details on the fiber parameters were presented in [35]. Taking the advantage of the fact that the soliton shift effect is peak power related, the center wavelength of the Raman-shifted soliton could be easily tuned simply by varying the pump power delivered to the EDFA connected with the PCF. It is worth noting that the Raman shifting process is fully coherent [35]. The effect of changing the peak power of pulses coupled into the PCF on the position of the Raman-shifted soliton is depicted in Fig. 3(b).
Fig. 3. (a) The peak power and average power of pulses launched into the PCF as a function of pump power delivered to the EDFA; (b) the Raman-shift-related evolution of the optical spectrum measured at the output of the highly nonlinear PCF recorded as a function of pulse peak power.
The EDFA was set to deliver 42 kW peak power pulses, which corresponded to shifting the soliton to a center wavelength of 1965nm. The pulses were subsequently chirped in a segment of a DCF (identical to the one used in the 1.56 µm section, −36 ps/nm/km at 1965nm), amplified in a thulium-doped fiber amplifier (TDFA), and re-compressed in a standard PM1550 fiber. At the maximum pump power delivered to the TDFA the Raman-shifted 2 µm pulses registered at the output of the common WDM reached a duration of 80 fs, an average output power of 300 mW, and a center wavelength of 1965nm. The optical spectra gathered at the output of the WDM along with autocorrelation traces of both pulses are shown in Fig. 4. The optical spectrum of the pulses amplified in the first section was centered at 1560 nm and showed signs of typical self-phase modulation-related broadening and modulation [36]. Water absorption peaks are clearly visible in the Raman-shifted part of the spectrum due to a free-space path inside the optical spectrum analyzer used for characterizing the emission. In the DFG experiment, the free space path was limited to ∼10 cm, which corresponds to ∼25% absorption of water vapor, considering the strongest peaks located between 1900nm and 1920nm.
Fig. 4. (a) Optical spectra of pulses taking part in the DFG process, Autocorrelation traces of the 1560 nm pulses - (b) and the 1960nm pulses – (c).
The optical path length of both fiber-based sections was designed to be nearly equal, therefore only slight tuning using the ODL was required to achieve coarse pulse overlap at the output of the WDM coupler. The combined 1.56 µm and 1.96 µm radiation was subsequently out-coupled from the common fiber onto an off-axis collimator and focused with an achromatic doublet lens (f = 75 mm) on a 3 mm-long orientation patterned gallium phosphide (OPGaP) crystal (BAE Systems) to generate mid-IR radiation via the DFG process. The choice of quasi-phase matched crystals with transparency in the mid-IR spectral range is limited to orientation-patterned semiconductors: GaP and GaAs. We have chosen OPGaP because of its high nonlinear coefficient (d = 70.6 pm∕V) and high thermal conductivity (110 W/mK) [37]. Recent research on the main competitor of GaP, namely GaAs, has revealed strong three-photon absorption (TPA) at ∼1.95 µm [22], which could influence the performance of the DFG process. To minimize temperature-related drifts the crystal was stabilized at 35 °C. The DFG setup based on common collimation optics for both interacting beams, connected with appropriate tuning of the ODL ensured nearly perfect spatial and temporal overlap of the pulses in the OPGaP crystal. As a result of the nonlinear DFG process, coherent mid-IR radiation was generated. The 3 dB bandwidth of the spectrum taking part in the nonlinear process was significantly broader than the theoretical acceptance bandwidth of the 3-mm-long crystal, therefore the center wavelength of the emission could be tuned easily over a broad range just by changing the poling period of the OPGaP crystal. Based on the GaP refractive index data [37] we estimate the group velocity mismatch (GVM) of the pump (1560 nm) and the signal (1950nm) wavelengths to be at the level of 133 fs/mm. The chosen 3 mm crystal length is a trade-off between the efficiency of the nonlinear process and the achievable bandwidth of the mid-IR idler. Nevertheless, thanks to the broadband input pulses the proposed DFG source was capable of covering a region between 1100 cm−1 and 1530 cm−1 (6535 nm – 9090 nm) simply by changing the QPM period of the OPGaP crystal. Optical spectra generated in the DFG process are plotted in Fig. 5 as a function of the OPGaP crystal poling period (registered with an FTIR spectrometer with 0.125 cm−1 resolution, Nicolet iS50). The spectra recorded between 1300 cm−1 and 1530 cm−1 show signs of water vapor absorption, due to a long free-space optical path between the crystal and the spectrometer. For a poling period equal to 61 µm the source was capable of delivering pulses with 5 mW average output power and a 3 dB bandwidth of 360 nm.
Fig. 5. The optical spectra of the Mid-IR pulses generated in the nonlinear DFG process plotted as a function of the OPGaP crystal poling period. The average output power of the generated radiation is plotted in yellow stars - right Y-axis.
3. Repetition rate and output power stabilization
DFG sources constructed in a common-seed configuration generate intrinsically fCEO-free pulses [38], therefore full control of the mid-IR comb is achieved by controlling only the repetition frequency (frep) and output power. Moreover, some comb-based gas sensing techniques, like continuous-filtering Vernier spectroscopy, require fast tuning of the frep with a ∼1 kHz amplitude [39]. In ML laser configurations based on bulk-optics this requirement is resolved by mounting an optical element on a piezo-ceramic transducer (PZT), which modulates the cavity length, and thus enables achieving and maintaining an appropriate pulse repetition rate [40]. In our setup the adjustment and stabilization of this parameter was accomplished in two steps. Initial frep tuning was realized by mounting the entire fiber-based ML laser resonator onto a custom-designed heating board enclosed in a 3D-printed enclosure. Although intrinsically slow, this method allowed for coarse adjustment of the frep [by 15.5 kHz, as shown in Fig. 6(a)] and limited the influence of ambient temperature drifts (by heating and stabilizing the resonator above the ambient temperature, e.g. at 30°C). Fast frep tuning was achieved by incorporating a 40 mm-long PZT stack epoxy-glued onto the active erbium-doped fiber, forming a simple fiber stretcher, whose frequency response is shown in Fig. 6(c). This element enables fast tuning of the frep in the range of 2.9 kHz and was used to lock the frep to a local oscillator (LO). As depicted on the schematic in Fig. 1, frep was monitored by a detector and mixed with the signal from a LO (signal generator, Tektronix, AFG3102) set to a frequency of ∼125.07 MHz. The differential signal is fed to a three-term controller (PID, model SRS SIM960) which produces an error signal for the piezo driver (PD, Thorlabs, MDT693B) controlling the fiber elongation in the ML laser resonator, and in turn allows for precise stabilization of the frep. To measure the frep drift we used an RF spectrum analyzer (Agilent, EXA N9010A) and a custom-made LabView software. The software monitored the frequency of the 56th beatnote of the ML laser (to increase the resolution of the measurement) at a 0.5 Hz rate.
Fig. 6. Results of the repetition frequency stabilization of the ML seed laser. (a) The RF spectrum of the fundamental beatnote registered for extreme values of resonator temperature and voltage applied to the PZT stretcher incorporated into the resonator. (b) The time dependent pulse repetition frequency with active temperature stabilization of the fiber resonator at 30°C (blue shaded) and with active stabilization to a LO (red shaded). (c) The frequency response of the in-house built fiber PZT stretcher for a RMS voltage of ∼3.5 V.
Figure 6(b) shows the results of the long-term stabilization of frep. Stabilizing the ML laser resonator temperature resulted in minimizing the frep fluctuations down to ∼100 Hz (0.8 ppm relative stability) within 2.5 hours after turning the laser on. The remaining, low-frequency variations were compensated with the LO-based PID feedback loop driving the PZT stretcher integrated into the oscillator. By combining the thermal and the PZT-based stabilization the long-term frep drift of the ML laser did not exceed 3 Hz during a 10 hour measurement (0.024 ppm), which is more than acceptable for applications in comb-based detection techniques requiring stable pulse repetition. The main limiting factor in this case was the stability of the LO used in the experiment as the frequency reference (1 ppm in 24 h). Using more sophisticated LO (e.g. GPS referenced) would further improve the frep stability.
Most laser-based gas-spectroscopy techniques require stable average output power and optical spectrum. This is also true for sensors using broadband comb sources. Since most of the source was constructed in an all-PM-fiber configuration, the majority of the vibration noise and long-term misalignment commonly present in bulk-optics-based systems was not affecting its overall performance and stability. Based on our observations, the main cause of output power instabilities was the drift of the temporal overlap between the pump and signal pulses during the DFG process. A change in the ambient temperature affects both the physical length of the fibers as well as their effective refractive index. Because both sections of the source are physically separated and have dissimilar pumping schemes (different amounts of optical power delivered to the active fibers), the accumulated variations are not correlated and thus directly influence the pulse overlap, which in turn manifests itself as variations in the generated spectrum and average output power. Recently, Silva de Oliveira et al. [31] proposed a method for stabilizing the pump-signal delay based on monitoring the residual intensity noise (RIN) in a certain frequency range. This technique allows for building a pulse overlap locking mechanism, which is non-complex and yields results superior to a straight-forward stabilization based on monitoring the output power. In our setup the electrical signal registered by a mid-IR detector with an appropriate bandwidth was bandpass filtered in the frequency region between 1 kHz to 300 kHz, amplified with a logarithmic amplifier (AD8310) and delivered to a servo controller (New Focus, LB1005) that drived a PZT controller (Thorlabs, MDT693B), forming a feed-back loop. The filtered and logarithmically amplified RIN component of the signal between 1 kHz and 100 kHz has a characteristic dip in the amplitude when the overlap between the pulses taking part in the DFG process is optimal, similarly to what was shown in [31] (corresponding to a maximum output power of the nonlinear process). An oscilloscope measurement showing the registered PZT ramp signal, the idler power and the associated RIN signal is plotted in Fig. 7.
Fig. 7. Integrated RIN spectral density in the range between 1 kHz and 100 kHz measured for the generated idler (left axis) and average idler output power (right axis) plotted in function of pump and signal pulse overlap (delay). Voltage ramp on the PZT element is plotted in blue. The 3 dB area of the RIN signal is shaded and corresponds to a delay offset of 4 fs.
As shown in Fig. 7, by applying a saw-tooth modulation to the PZT element acting as a tunable optical delay a clear maximum of idler output power and a sharp dip in the integrated RIN can be observed. Having a significant amplitude, this dip can be successfully exploited to stabilize the overlap between the pump and signal pulses in the OPGaP crystal, and thus produce stable average output power. It is worth noting that the 3 dB width of the RIN dip corresponds to only a 4 fs pump-signal delay, which is more than 4x narrower compared to the results presented in [31]. We believe this improvement is mainly caused by the all-in-fiber configuration of our source, compared to the free space source in ref. [31]. The robustness of the fiber-based structure results in a lower overall noise, thus the RIN minimum can be observed in a narrower region of the pulse delay. In our experiment the 4 fs offset result in a ∼0.4% change in the average idler output power, therefore it is obvious that the method for stabilizing the DFG process presented in [31] is far superior compared to a simple amplitude lock. Subsequently the RIN minimum signal was utilized to lock the offset between the pulses taking part in the DFG process. The output power and the generated spectra registered for 60 minutes for a stabilized and un-stabilized case are plotted in Fig. 8.
Fig. 8. 60 minute heatmaps calculated from the DFG output spectra gathered every 60 seconds, with (a) the active stabilization turned OFF, and (b) ON. Panel (c) shows the output power stability for both cases measured as a function of time.
Simultaneous monitoring of time-dependent evolution of the optical spectrum and the average output power required splitting the DFG beam using a beamsplitter, and directing it separately to a mid-IR detector and the FT-IR spectrometer, therefore the graph in Fig. 8(c) shows a maximum average output power of ∼2.3 mW during the stabilization process (in our experiment a 50/50 splitter was used). Although the measurements were taken after an initial 10 minute burn-in of the fiber-based source the output power and the generated optical spectra exhibited a significant drift over a 60 minute-long measurement in a free-running case [plotted in Fig. 8(a) and highlighted in red in Fig. 8(c)]. After enabling the stabilization loop both parameters remained stabilized over the entire period of the measurement. Standard deviation of the generated output power was 0.4 mW and 0.019 mW for the free-running and the stabilized case, respectively, which shows a 21 fold increase in the stability. The achieved stability outperforms the previously reported long-wave and short-wave mid-infrared DFG sources. In [30] the output power dropped from 6.5 to 6.2 mW after 60 minutes. The authors in [41] report a short-wavelength (3 µm) DFG with measured 3.2% of root mean square (RMS) stability, compared to 0.8% in our system. Significant improvement in pulse overlap resulted in generating a stable spectrum over the entire 60 minute measurement period [highlighted in green in Fig. 8(c)], which confirms that the applied stabilization method can be successfully used in DFG-based mid-IR comb sources for applications in gas spectroscopy. We observed that the main limiting factor for the stabilization technique used in our experiment was the high noise of the mid-IR detector (Thorlabs PDAVJ10), which can be clearly seen on the measurements in Fig. 8(c). Moreover, the overall stability can be improved by implementing active stabilization of the output power in each amplification stage (at a cost of increased complexity of the system).
4. Absorption spectroscopy
To demonstrate the applicability of our DFG source for spectroscopic measurements we measured the absorption spectrum of the v1 band of N2O around 7.8 µm. The beam of the DFG source with the 61 µm OPGaP crystal was transmitted through a 10 cm-long absorption cell filled with 0.75% of N2O in N2 at 760 Torr and measured with a Fourier transform spectrometer (Thermo Scientific Nicolet iS50, 0.125 cm−1 resolution). The resulting absorption spectrum of the gas sample in the range between 1230 cm−1 and 1330 cm−1 (7520–8130 nm) is plotted in black in Fig. 9, along with a fit of a model based on the parameters from the HITRAN database [42], shown in red. The experimental spectrum was first normalized to a background spectrum measured with the cell filled with pure nitrogen. The fit was performed using a linear method based on the cepstral analysis [43], which also allowed removing the slowly-varying baseline remaining after normalization. The only fitting parameter was the concentration, which was found to be 0.75%, and the general agreement between the model and the spectrum is very good. The structure in the residuum, shown in the lower panel of Fig. 9, might be caused by inaccuracies in the line positions and broadening parameters in the HITRAN database, as well as by minor distortions in the spectrum. In particular, the spikes in the spectrum, visible e.g. at 1317 cm−1 and 1319 cm−1, are caused by large water absorption in the background spectrum originating from the free space path outside the cell.
Fig. 9. Absorption spectrum of 0.75% N2O in N2 at 760 Torr (upper panel, black curve). A fit based on HITRAN parameters is plotted in red, and the residuum is shown in the lower panel.
5. Conclusions
In this work a simple configuration of a fully-fiberized DFG-based fCEO-free mid-IR optical frequency comb source with 125 MHz repetition rate was presented. The source is capable of delivering up to 5 mW of coherent radiation tunable in the range between 6.5 µm – 9 µm simply by changing the OPGaP crystal in which the nonlinear process occurs. This is also the first demonstration of a mid-IR all-fiber comb source with active stabilization of the repetition frequency and the temporal pulse overlap during the nonlinear frequency conversion process. The delay stabilization resulted in a significant increase of the output pulse stability (by a factor of 21) and ensured stable idler beam spectrum during a 1 hour measurement period. The measurement of the offset-related RIN amplitude integrated between 1 kHz and 100 kHz shows a 4-fold decrease of the 3 dB width compared to ref. [31], which shows that the all-in-fiber configuration of our source is less susceptible to noise sources such as mechanical vibrations, temperature drifts, etc. The feasibility of the presented source for gas detection applications was experimentally verified by measuring absorption spectra of nitrous oxide (N2O), which is a greenhouse gas with significant global warming potential. The overall robustness and long-term stability of the constructed source paves the way to out-of-lab applications, e.g. in environmental multispecies monitoring of greenhouse gas emissions, where broadband sources can be an interesting alternative to using separate single-mode sources, each targeting a different molecule transition.
Funding
Fundacja na rzecz Nauki Polskiej (First TEAM/2017-4/39); Knut och Alice Wallenbergs Stiftelse (KAW 2015.0159); Politechnika Wroclawska.
References
1. K. Krzempek, R. Lewicki, L. Nähle, M. Fischer, J. Koeth, S. Belahsene, Y. Rouillard, L. Worschech, and F. K. Tittel, “Continuous wave, distributed feedback diode laser based sensor for trace-gas detection of ethane,” Appl. Phys. B 106(2), 251–255 (2012). [CrossRef]
2. U. Willer, M. Saraji, A. Khorsandi, P. Geiser, and W. Schade, “Near-and mid-infrared laser monitoring of industrial processes, environment and security applications,” Opt. Laser Eng. 44(7), 699–710 (2006). [CrossRef]
3. L. Tao, K. Sun, D. J. Miller, D. Pan, L. M. Golston, and M. A. Zondlo, “Low-power, open-path mobile sensing platform for high-resolution measurements of greenhouse gases and air pollutants,” Appl. Phys. B 119(1), 153–164 (2015). [CrossRef]
4. G. Villares, A. Hugi, S. Blaser, and J. Faist, “Dual-comb spectroscopy based on quantum-cascade-laser frequency combs, Nature communications,” Nat. Commun. 5(1), 5192 (2014). [CrossRef]
5. L. A. Sterczewski, J. Westberg, and G. Wysocki, “Computational coherent averaging for free-running dual-comb spectroscopy,” Opt. Express 27(17), 23875–23893 (2019). [CrossRef]
6. J. Hult, R. S. Watt, and C. F. Kaminski, “High bandwidth absorption spectroscopy with a dispersed supercontinuum source,” Opt. Express 15(18), 11385–11395 (2007). [CrossRef]
7. K. Eslami Jahromi, Q. Pan, L. Høgstedt, S. M. M. Friis, A. Khodabakhsh, P. M. Moselund, and F. J. M. Harren, “Mid-infrared supercontinuum-based upconversion detection for trace gas sensing,” Opt. Express 27(17), 24469–24480 (2019). [CrossRef]
8. A. G. Griffith, R. K. W. Lau, J. Cardenas, Y. Okawachi, A. Mohanty, R. Fain, Y. H. D. Lee, M. Yu, C. T. Phare, C. B. Poitras, A. L. Gaeta, and M. Lipson, “Silicon-chip mid-infrared frequency comb generation,” Nat. Commun. 6(1), 6299 (2015). [CrossRef]
9. M. Yu, Y. Okawachi, A. G. Griffith, M. Lipson, and A. L. Gaeta, “Mode-locked mid-infrared frequency combs in a silicon microresonator,” Optica 3(8), 854–860 (2016). [CrossRef]
10. M. Vainio and L. Halonen, “Mid-infrared optical parametric oscillators and frequency combs for molecular spectroscopy,” Phys. Chem. Chem. Phys. 18(6), 4266–4294 (2016). [CrossRef]
11. A. Schliesser, N. Picque, and T. W. Hänsch, “Mid-infrared frequency combs,” Nat. Photonics 6(7), 440–449 (2012). [CrossRef]
12. A. Ruehl, A. Gambetta, I. Hartl, M. E. Fermann, K. S. E. Eikema, and M. Marangoni, “Widely-tunable mid-infrared frequency comb source based on difference frequency generation,” Opt. Lett. 37(12), 2232–2234 (2012). [CrossRef]
13. F. Adler, K. C. Cossel, M. J. Thorpe, I. Hartl, M. E. Fermann, and J. Ye, “Phase-stabilized, 1.5 W frequency comb at 2.8–4.8 µm,” Opt. Lett. 34(9), 1330–1332 (2009). [CrossRef]
14. D. Sánchez, M. Hemmer, M. Baudisch, K. Zawilski, P. Schunemann, H. Hoogland, R. Holzwarth, and J. Biegert, “Broadband mid-IR frequency comb with CdSiP2 and AgGaS2 from an Er,Tm:Ho fiber laser,” Opt. Lett. 39(24), 6883–6886 (2014). [CrossRef]
15. K. Krzempek, G. Sobon, and K. M. Abramski, “DFG-based mid-IR generation using a compact dual-wavelength all-fiber amplifier for laser spectroscopy applications,” Opt. Express 21(17), 20023–20031 (2013). [CrossRef]
16. G. Ycas, F. R. Giorgetta, E. Baumann, I. Coddington, D. Herman, S. A. Diddams, and N. R. Newbury, “High-coherence mid-infrared dual-comb spectroscopy spanning 2.6 to 5.2 µm,” Nat. Photonics 12(4), 202–208 (2018). [CrossRef]
17. F. Adler, P. Masłowski, A. Foltynowicz, K. C. Cossel, T. C. Briles, I. Hartl, and J. Ye, “Mid-infrared Fourier transform spectroscopy with a broadband frequency comb,” Opt. Express 18(21), 21861–21872 (2010). [CrossRef]
18. L. Nugent-Glandorf, T. Neely, F. Adler, A. J. Fleisher, K. C. Cossel, B. Bjork, T. Dinneen, J. Ye, and S. A. Diddams, “Mid-infrared virtually imaged phased array spectrometer for rapid and broadband trace gas detection,” Opt. Lett. 37(15), 3285–3287 (2012). [CrossRef]
19. A. Khodabakhsh, V. Ramaiah-Badarla, L. Rutkowski, A. C. Johansson, K. F. Lee, J. Jiang, C. Mohr, M. E. Fermann, and A. Foltynowicz, “Fourier transform and Vernier spectroscopy using an optical frequency comb at 3–5.4 µm,” Opt. Lett. 41(11), 2541–2544 (2016). [CrossRef]
20. Z. Chen, T. W. Hänsch, and N. Picqué, “Mid-infrared feed-forward dual-comb spectroscopy,” Proc. Natl. Acad. Sci. U. S. A. 116(9), 3454–3459 (2019). [CrossRef]
21. N. Leindecker, A. Marandi, R. L. Byer, K. L. Vodopyanov, J. Jiang, I. Hartl, M. Fermann, and P. G. Schunemann, “Octave-spanning ultrafast OPO with 2.6-6.1 µm instantaneous bandwidth pumped by femtosecond Tm-fiber laser,” Opt. Express 20(7), 7046–7053 (2012). [CrossRef]
22. O. H. Heckl, B. J. Bjork, G. Winkler, P. B. Changala, B. Spaun, G. Porat, T. Q. Bui, K. F. Lee, J. Jiang, M. E. Fermann, P. G. Schunemann, and J. Ye, “Three-photon absorption in optical parametric oscillators based on OP-GaAs,” Opt. Lett. 41(22), 5405–5408 (2016). [CrossRef]
23. L. Jin, V. Sonnenschein, M. Yamanaka, H. Tomita, T. Iguchi, A. Sato, K. Nozawa, K. Yoshida, S.-I. Ninomiya, and N. Nishizawa, “3.1–5.2 µm Coherent MIR Frequency Comb Based on Yb-Doped Fiber Laser,” IEEE J. Sel. Top. Quantum Electron. 24(3), 1–7 (2018). [CrossRef]
24. D. L. Maser, G. Ycas, W. I. Depetri, F. C. Cruz, and S. A. Diddams, “Coherent frequency combs for spectroscopy across the 3–5 µm region,” Appl. Phys. B 123(5), 142 (2017). [CrossRef]
25. C. R. Phillips, J. Jiang, C. Mohr, A. C. Lin, C. Langrock, M. Snure, D. Bliss, M. Zhu, I. Hartl, J. S. Harris, M. E. Fermann, and M. M. Fejer, “Widely tunable midinfrared difference frequency generation in orientation-patterned GaAs pumped with a femtosecond Tm-fiber system,” Opt. Lett. 37(14), 2928–2930 (2012). [CrossRef]
26. M. Beutler, I. Rimke, E. Büttner, P. Farinello, A. Agnesi, V. Badikov, D. Badikov, and V. Petrov, “Difference-frequency generation of ultrashort pulses in the mid-IR using Yb-fiber pump systems and AgGaSe2,” Opt. Express 23(3), 2730–2736 (2015). [CrossRef]
27. C. Gaida, M. Gebhardt, T. Heuermann, F. Stutzki, C. Jauregui, J. Antonio-Lopez, A. Schülzgen, R. Amezcua-Correa, A. Tünnermann, I. Pupeza, and J. Limpert, “Watt-scale super-octave mid-infrared intrapulse difference frequency generation,” Light: Sci. Appl. 7(1), 94 (2018). [CrossRef]
28. A. Gambetta, N. Coluccelli, M. Cassinerio, D. Gatti, P. Laporta, G. Galzerano, and M. Marangoni, “Milliwatt-level frequency combs in the 8–14 µm range via difference frequency generation from an Er:fiber oscillator,” Opt. Lett. 38(7), 1155–1157 (2013). [CrossRef]
29. G. Soboń, T. Martynkien, P. Mergo, L. Rutkowski, and A. Foltynowicz, “High-power frequency comb source tunable from 2.7 to 4.2 µm based on difference frequency generation pumped by an Yb-doped fiber laser,” Opt. Lett. 42(9), 1748–1751 (2017). [CrossRef]
30. J. Sotor, T. Martynkien, P. G. Schunemann, P. Mergo, L. Rutkowski, and G. Soboń, “All-fiber mid-infrared source tunable from 6 to 9 µm based on difference frequency generation in OP-GaP crystal,” Opt. Express 26(9), 11756–11763 (2018). [CrossRef]
31. V. Silva de Oliveira, A. Ruehl, P. Masłowski, and I. Hartl, “Intensity Noise Optimization of a Mid-Infrared Frequency Comb Difference Frequency Generation Source,” arXiv:1904.02611 [physics.optics] (2019).
32. A. Krajewska, I. Pasternak, G. Sobon, J. Sotor, A. Przewloka, T. Ciuk, J. Sobieski, J. Grzonka, K. M. Abramski, and W. Strupinski, “Fabrication and applications of multi-layer graphene stack on transparent polymer,” Appl. Phys. Lett. 110(4), 041901 (2017). [CrossRef]
33. J. W. Nicholson, A. D. Yablon, P. S. Westbrook, K. S. Feder, and M. F. Yan, “High power, single mode, all-fiber source of femtosecond pulses at 1550 nm and its use in supercontinuum generation,” Opt. Express 12(13), 3025–3034 (2004). [CrossRef]
34. J. Takayanagi, T. Sugiura, M. Yoshida, and N. Nishizawa, “1.0-1.7-µm Wavelength-Tunable Ultrashort-Pulse Generation Using Femtosecond Yb-Doped Fiber Laser and Photonic Crystal Fiber,” IEEE Photonics Technol. Lett. 18(21), 2284–2286 (2006). [CrossRef]
35. G. Soboń, T. Martynkien, D. Tomaszewska, K. Tarnowski, P. Mergo, and J. Sotor, “All-in-fiber amplification and compression of coherent frequency-shifted solitons tunable in the 1800–2000 nm range,” Photonics Res. 6(5), 368–372 (2018). [CrossRef]
36. R. H. Stolen and C. Lin, “Self-phase-modulation in silica optical fibers,” Phys. Rev. A 17(4), 1448–1453 (1978). [CrossRef]
37. 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]
38. M. Zimmermann, C. Gohle, R. Holzwarth, T. Udem, and T. W. Hänsch, “Optical clockwork with an offset-free difference-frequency comb: accuracy of sum- and difference-frequency generation,” Opt. Lett. 29(3), 310–312 (2004). [CrossRef]
39. L. Rutkowski and J. Morville, “Broadband cavity-enhanced molecular spectra from Vernier filtering of a complete frequency comb,” Opt. Lett. 39(23), 6664–6667 (2014). [CrossRef]
40. S. N. Bagayev, S. V. Chepurov, V. M. Klementyev, S. A. Kuznetsov, V. S. Pivtsov, V. V. Pokasov, and V. F. Zakharyash, “A femtosecond self-mode-locked Ti: sapphire laser with high stability of pulse-repetition frequency and its applications,” Appl. Phys. B 70(3), 375–378 (2000). [CrossRef]
41. J. C. Casals, S. Parsa, S. Chaitanya Kumar, K. Devi, P. G. Schunemann, and M. Ebrahim-Zadeh, “Picosecond difference-frequency-generation in orientation-patterned gallium phosphide,” Opt. Express 25(16), 19595–19602 (2017). [CrossRef]
42. I. E. Gordon, L. S. Rothman, C. Hill, R. V. Kochanov, Y. Tan, P. F. Bernath, M. Birk, V. Boudon, A. Campargue, K. V. Chance, B. J. Drouin, J.-M. Flaud, R. R. Gamache, J. T. Hodges, D. Jacquemart, V. I. Perevalov, A. Perrin, K. P. Shine, M.-A. H. Smith, P. J. Tennyson, G. C. Toon, H. Tran, V. G. Tyuterev, A. Barbe, A. G. Csaszar, V. M. Devi, T. Furtenbacher, J. J. Harrison, J.-M. Hartmannn, A. Jolly, T. J. Johnson, T. Karman, I. Kleiner, A. A. Kyuberis, J. Loos, O. M. Lyulin, S. T. Massie, S. N. Mikhailenko, N. Moazzen-Ahmadi, H. S. P. Müller, O. V. Naumenko, A. V. Nikitin, O. L. Polyansky, M. Rey, M. Rotger, S. W. Sharpe, K. Sung, E. Starikova, S. A. Tashkun, J. Vander Auwera, G. Wagner, J. Wilzewski, P. Wcislo, S. Yu, and E. J. Zak, “The HITRAN2016 molecular spectroscopic database,” J. Quant. Spectrosc. Radiat. Transfer 203, 3–69 (2017). [CrossRef]
43. R. K. Cole, A. Makowiecki, N. Hoghooghi, and G. B. Rieker, “Baseline-free Quantitative Absorption Spectroscopy Based on Cepstral Analysis,” arXiv:1906.11349 (2019).
