Open Access
Add to BibSonomy
Abstract
Pulsed nonlinear-optical upconversion is used for mid-infrared signal detection. A setup for both mid-infrared generation and upconversion based on a single pump laser enables sensitive light detection and is utilized for gas spectroscopy. With the demonstrated pulsed setup, quantum efficiencies above 80 % for the upconversion of a Gaussian beam signal and 25 % for the upconversion of backscattered radiation are achieved and in agreement with theoretical predictions. Combined with efficient background suppression due to spectral and temporal gating, this results in highly sensitive detection of the infrared signals. As a demonstration of application, the presented system is used for methane sensing in an open path geometry, highlighting the potential for stand-off leak detection with a concentration resolution better than 1.5 ppm·m.
© 2017 Optical Society of America
1. Introduction
The spectroscopy of molecular rotational-vibrational transitions in the infrared spectral range offers a toolbox for the analysis of a wide range of substances. One important application is the stand-off detection of infrared-active gases. A gas of particular interest is methane (CH4), since it is one of the top three primary energy sources worldwide [1], and leakages are not only a safety concern, but also largely contribute to the global greenhouse effect [2]. Therefore, methane leak detection is an important issue, often over large areas and distances.
For stand-off CH4 sensing from a distance, infrared laser spectroscopy is commonly employed [3–5]. Existing systems are usually operated in the near-infrared (NIR) range around the 1.67-m overtone band of CH4 [3–6]. The strong fundamental C-H-band of hydrocarbons in the mid-infrared (MIR) range allows for a higher sensitivity [6]. However, MIR detector materials like mercury cadmium telluride or indium antimonide have several drawbacks compared to materials for the near-infrared and visible range, such as silicon or gallium arsenide. Because of the smaller energy band-gap, MIR detectors are severely limited in sensitivity by thermal noise [7]. The detectors usually need to be extensively cooled and offer smaller and less uniform sensitive areas than their silicon counterparts. Additionally, visible and NIR detectors have further advantages in terms of detection bandwidth and cost.
A method to bridge the gap between these spectral ranges is nonlinear-optical upconversion [8]. By mixing an MIR signal with a pump laser in a nonlinear optical medium, the signal can be transferred to shorter wavelength ranges, where detectors with more suitable properties are available. Over the last few years, significant advances in upconversion technology have demonstrated its feasibility for low-noise infrared detection, e.g. in astronomy, spectroscopy and imaging [9–13]. Recent setups for the upconversion of thermal sources are often continuously pumped, in order not to reduce the effective conversion efficiency for continuous signals by the pulse duty cycle. For efficient continuous-wave upconversion at low pump input power levels, intracavity upconversion is employed [13].
This work investigates for sensitive detection of a laser-based MIR signal the application of a pulsed setup, in which pump pulses are used for both the generation of MIR light in one optical branch and the subsequent upconversion for silicon-based detection in the second branch. This coupled-pulse downconversion-upconversion approach has been demonstrated in [14–16]. The method does not only allow for efficient upconversion of the pulsed signal, it also intrinsically provides a suppression of background noise analogous to spectral filtering and time gating techniques, as they are often used in spectroscopy [17,18]. Here, a system is presented, achieving high total upconversion efficiencies. A quantitative efficiency analysis through measurement and calculation is given. The implications for the sensitivity of MIR detection are discussed.
The beneficial detection properties allow the setup to be used for the sensing of methane in an open-path backscattering geometry. The system reaches a low limit of detection at a path concentration below 1.5 ppm·m of CH4, achieved with a simple differential-wavelength measurement scheme. The results show promising potential for sensitive measurements over long distances, on weakly scattering targets and against increased thermal radiation background.
2. Nonlinear-optical three-wave mixing
Both the generation and the upconversion of mid-infrared radiation in the presented system are based on the mechanism of three wave mixing in a nonlinear-optical crystal driven by the same pulsed pump laser. The mid-infrared pulses with the optical frequency ωm originate from difference frequency generation (DFG):
Herein, ωp is the frequency of the pump light and ωi that of an injected second wave also amplified in the interaction. By controlling ωi, spectrally well-defined, narrow-bandwidth radiation is generated at ωm. This so-called seeded optical parametric generation (OPG) and amplification (OPA) process is described in detail in precious work [19,20]. For detection of the mid-infrared signal using silicon detectors, the radiation is upconverted by sum frequency generation (SFG), the inverse process of the DFG. Here, energy is transferred from the waves at ωp and ωm to a generated fourth wave with the sum frequencyThe key figure-of-merit for the SFG process is the upconversion efficiency ηSFG, defined as the ratio of the generated flux of upconverted photons Φs to the incident mid-IR photon flux Φm,
The upconversion efficiency is in general determined by the medium’s nonlinear coefficient deff and length L, the electrical field amplitude Ep of the pump wave and the phase mismatch parameter ∆k of the interacting waves, which, in the case of quasi-phase-matching (QPM), is given by
The ni denote the corresponding refractive indices at the wave frequencies and c0 the vacuum speed of light. The poling period Λ of the medium contributes to ∆k as 2π/Λ [21]. Depending on the incident photon flux ratios of infrared signal and pump laser waves, significant depletion of the pump wave can occur. A solution of the coupled differential field equations for the three waves taking into account the depletion leads to the following expression for the quantum efficiency [22]:
Herein, sn () denotes the Jacobi elliptic function. The notation includes
where Ep,0 and Em,0 are the electrical field amplitudes of pump and MIR wave at the surface of the nonlinear medium, andWhen the pump depletion can be neglected, e.g. for Φm ≪ Φp, the expression for the quantum efficiency is simplified to [23]
For the SFG of pulsed MIR and pump light, the total conversion efficiency for a mid-infrared signal is strongly influenced by spatial and temporal shape of both pulses. For the calculation of the total conversion efficiency, the efficiency distribution ηSFG (x, y, t) is evaluated over the cross section of the crystal, where z is assumed to be the direction of collinear propagation of the light waves, and then weighted with the overlap of the MIR pulse shape. Influences of the group delay dispersion in the crystal are neglected here, which is a valid assumption for pulse durations in the nanosecond range and crystal sizes of few centimeters. The approach can account for varying spatial and temporal distributions. However, it disregards differing propagation angles and diffraction, and thus is not suitable for the treatment of noncollinear or tightly focused beams.
3. Generator-converter system
The system for pulsed generation of mid-infrared radiation and sub-sequent upconversion of the signal is shown schematically in Fig. 1. Both conversion processes are driven by the same pulsed laser. The laser (CryLas DSS 1064-Q) emits pulses with a center wavelength of 1064.4 nm and a temporal half-max duration of 700 ps. It is driven at a repetition rate of 1100 Hz with a pulse energy of 50 µJ. The laser pulses are split in a polarizing beamsplitter into two paths. The split ratio is controlled by rotation of a half-wave plate in front of the beamsplitter. A second half-wave plate rotates the transmitted fraction back to vertical polarization. Pump pulses of about 16 µJ energy are directed towards the MIR generation. In the generation branch, the pump light is combined with the collimated, continuous-wave beam of a tunable diode laser (Eblana EP1550-DM-B) through a dichroic mirror for the purpose of injection seeding of the OPG/OPA [19]. The combined beams are focused by a 50-mm-focal-length lens to the center of a first crystal of periodically-poled lithium niobate (PPLN1). The crystal with a length of 17 mm and a poled channel aperture of 1×1 mm2 is set for quasi-phase-matched OPG of 1551 and 3390 nm with a poling period of 30.5 µm and a temperature of 43.90 °C in a thermally stabilized mount. The generated mid-infrared difference-frequency radiation with wavelengths around 3390 nm is collimated with a 30-mm-focal-length germanium (Ge) lens that simultaneously acts as a beam blocker for the two shorter wavelength waves. The OPG/OPA yields an MIR beam with a fundamental gaussian mode of 1.1 mm 1/e2-intensity radius, as shown in Fig. 2(a), and pulse energies of 1.4 µJ.
Fig. 1 Basic scheme of the coupled-pulse remote sensing setup in a backscattering configuration. TDL: tunable diode laser; FC: fiber collimator; PL: pulse laser; HWP: half-wave plate; PBS: polarizing beamsplitter; DM: dichroic mirror; PPLN1/2: periodically poled lithium niobate crystals; ST: scattering target; DL: delay line; BPF: band pass filter set.
Fig. 2 a) Beam profile of the generated MIR radiation at 3390 nm after collimation. b) Temporal shape of the pulses. c) Radii of the MIR signal and the pump beams as they are superimposed for upconversion. The crystal dimensions are marked by the turquoise bar.
In the upconversion branch, the pump light passes through a fixed-mirror delay line that is set to match the optical path lengths such that the mid-infrared and the pump pulse coincide optimally in the second crystal (PPLN2, 20 mm length, 2×2 mm2 aperture for a poled channel). Both crystals are oriented such that all light is extraordinarily polarized. MIR and pump light are both focused towards the crystal center and superimposed through a second dichroic mirror. Phase matching conditions are set for SFG from 1064.4 nm and 3390 nm to 810 nm at a poling period of 22.34 µm and a temperature of 115.56 °C. The QPM provides an acceptance window FWHM of 10 nm in the MIR. A combination of filters with a transmittance band from 750 to 850 nm is placed behind the upconversion crystal to block the pump light.
The temporal shape of the pulses is measured with a fast indium gallium arsenide photodiode (New Focus 1544-A-50) and a high-bandwidth oscilloscope (LeCroy WaveMaster 8300A, 3 GHz). The traces are shown in Fig. 2(b). The pulses at the pump and SFG wavelengths can be detected directly, the temporal shape of the generated MIR pulse is determined by analyzing the correlated pulse at the seeding wavelength. Measured FWHM pulse durations are 700 ps for the pump pulse, 360 ps for the MIR and 310 ps for the SFG pulse. The reduced MIR pulse duration with respect to the pump pulse is due to the nonlinear pump power dependency of the generation. In consequence, the MIR pulse is fully enveloped by the pump pulse for the upconversion. The duration of the SFG pulse is thus mainly determined by the MIR pulse shape.
For the determination of the collinear upconversion efficiency, the generated MIR radiation is directly guided to the dichroic combiner mirror and collimated to a beam waist of 400 µm radius inside the upconversion crystal. The high peak power of the pump pulses allows for a large diameter of the pump beam, while maintaining intensities that lead to high upconversion effinciency. The pump beam waist in the crystal is therefore set with a 750-m radius such that it fully envelops the MIR signal. The beam profiles are measured with a pyroelectric array detector (Ophir PyroCam IIIHR). As Fig. 2(c) shows, the Rayleigh length greatly exceeds the crystal length for both beams, leading to a very good approximation of plane wave interaction. The upconversion efficiency is determined from average power measurements of the MIR and the SFG radiation with a thermopile sensor (Thorlabs S401C). For an unattenuated MIR branch with a static seed wavelength, the average incident power is measured and compared to the generated sum frequency power. The experimental upconversion efficiency is calculated as the ratio of the corresponding photon fluxes according to Eq. (3). For a comparison of the determined values against the theoretical prediction, a series of efficiency measurements is conducted for different pump powers impinging onto the SFG crystal. Figure 3(a) shows the results compared to the theoretical efficiency curves for full calculation according to Eq. (5) as well as for the case of neglected pump depletion as in Eq. (8). Good agreement is found for the experimental data with the calculation including depletion of the pump wave. A maximum total efficiency of (82 ± 1.5) % for the upconversion of the collimated MIR pulse is determined. The increasing deviation of the measured efficiency towards higher pump pulse energies might indicate an increasing impact of parasitic additional processes that set on at higher pump intensities. In the following, the MIR beam is projected on a gold-coated diffuse scattering target with a spot size of 2 mm diameter. Scattered radiation is imaged to the upconversion crystal center with a 75-mm-focal-length Ge lens positioned in a distance of 470 mm from the target. The power level of the backscattered radiation collected by the lens is too low for detection on a thermal sensor. Instead, a scientific grade silicon CMOS camera (Andor ZYLA 4.2) is used to measure the SFG light. With the known quantum efficiency and gain factor of the camera, the SFG power is calculated. An adjustable aperture in front of the collection lens is employed to determine the effective area of the lens. Figure 3(b) shows a sketch of the experimental configuration. A sensitive lens area of 7.5 mm diameter is found. This implies an effective field of view of the upconversion system of 5.4° full angle, corresponding to a numerical aperture of 0.05. From the known scattering distribution of the target, the fraction of light imaged to the converter is calculated. With the MIR beam attenuated to energies of 45.5 nJ per pulse before scattering, 7.3 pJ of scattered pulse energy are collected for upconversion. At these signal energies, depletion of the pump wave is fully negligible. A camera exposure of 33 pulses in 30 ms is evaluated. The measurement yields a calculated upconversion efficiency of η = (25 ± 3) % for the scattered MIR signal entering the aperture. The reduction of quantum efficiency compared to that of the collinear case is caused primarily by the increasing phase mismatch for light incident under larger angles, described in detail in [24], and by wavefront distortion and partial depolarization from the scattering on the target.
Fig. 3 a) Measured upconversion efficiency compared to theoretical calculations with neglected (dashed) and included (solid) pump wave depletion. b) Optical setup for the characterization of upconversion efficiency in the backscattering configuration.
A second exposure taken with a blocked MIR branch shows the background generated from the pump-illuminated SFG crystal to be very weak. In 20 s of dark exposure time, an average dark count rate of 5.7 counts per pixel and second is measured in the evaluated camera region. The background shows a homogenous intensity distribution with no observable angular preference. This suggests a background origin from upconverted broadband thermal radiation or scattered pump light rather than parasitic nonlinear-optical processes in the observed spectral window. Compared to a measurement of the backscattered MIR pulses, a signal-to-background ratio of 46 000 is determined even for the attenuated MIR imaged from the distance of 470 mm. This result is thanks to the combination of high upconversion quantum efficiency and the low-noise detection on a silicon device, as well as the additional suppression of thermal background by the coupled-pulse upconversion method. In comparison, detecting the same backscattered signal, corresponding to an average power of 8 nW, on an ungated MIR detector sensitive between 3 and 5 µm, the average signal-to-background ratio would be limited to a value of 6 just by a 300 K blackbody background in the same numerical aperture. The high signal-to-background ratio means that the system operates far from shot-noise limitation and allows for measurements covering significantly larger distances or weakly scattering target surfaces. Already with the MIR pulse energies of 1.4 µJ generated here, corresponding to 1.5 mW of average power, a ratio of 500 between the signal and measured background would still be achieveable for a distance of 17 m on a target with a reflectivity of 0.2.
4. Methane sensing
To characterize the system as an instrument for gas sensing, a transmission cell with 10 cm absorption path length and CaF2 windows is introduced into the MIR branch before the scattering target, as shown in Fig. 4(a). A calibrated mass flow controller (MFC, HovaCAL D922-SP) is used to dilute a sample gas of 1 % CH4 in N2 to adjustable concentrations.
Fig. 4 a) Setup configuration for the characterization of the system as a CH4 sensor. b) Measured transmission spectrum of 1000 ppm·m of CH4 in N2, compared to a simulation based on HITRAN data. c) Concentration series set in the sample gas cell by the MFC and transmission ratio measured by the DFG-SFG system. The inset shows a zoom on the transmission axis for the last four concentration levels.
The seed diode laser is modulated with a sawtooth current, leading to a wavelength sweep of the MIR radiation in a tuning range between 3390.5 and 3393.5 nm. The camera is replaced by a single unbiased silicon photodiode (Centronic OSD60-5T), the signal of which is amplified and fed to a DAQ-card. For a cell purged with pure N2, the SFG signal of a sweep is recorded as a reference. A path concentration of 1000 ppm·m CH4 is then set in the cell. Divided by the reference, a CH4 transmission spectrum is obtained with a scan rate of one spectrum per second. In Fig. 4(b), the result is shown to be in good agreement with a simulated spectrum based on HITRAN data [25]. The simulation takes into account the spectral instrument function determined by the bandwidth of the MIR signal. Best matching is achieved with a simulated spectral bandwidth of 70 pm or 1.8 GHz, close to the Fourier limited value of 1.4 GHz corresponding to the measured FWHM duration of the MIR pulses.
Since the sampling rate of the spectrum is limited by the repetition rate of the pulse laser, a differential-wavelength measurement scheme is favored over a continuous spectral scan in order to achieve a higher measurement rate. The seed laser diode is hence square-modulated with a frequency of 10 Hz such that alternating MIR wavelengths of λon = 3391.62 nm on the absorption line center and λoff = 3390.65 nm in a region of weak CH4-absorption are emitted. The seeding wavelength repeatability is measured to be better than 0.5 pm with a wavemeter (HighFinesse WS7-30 IR2) over the measurement time. The deviations are negligibly small compared to the linewidth of the pulsed pump laser and hence the MIR pulse spectral width. With a certain hold-off time during the tuning, the measured signal at each wavelength is the result of averaging over 33 pulses. In this measuring mode, a series of concentrations of CH4 in N2 is set with the MFC, corresponding to path concentrations ranging from 1000 down to 1.5 ppm·m. The measured transmission ratio between the two wavelengths is shown in correlation to the set path concentration in Fig. 4(c). Minor signal drift in the range of 0.2 % over 1000 s is here compensated by a normalizing function derived from the averaged values for the nitrogen-purged periods over the measurement time. This drift is due to long-term fluctuations in the pump laser power, which can be monitored for intrinsic drift compensation as a possible improvement of the setup. The measurement resolves the steps well over the full dynamic range. The absorption caused by a path concentration of 1.5 ppm·m can still be detected at a 4-σ-margin. From the baseline noise, a noise equivalent absorptance of 2 · 10−4 Hz−1/2 is deduced, corresponding to a path concentration detection limit of 0.2 ppm·m under laboratory conditions. Compared to the path concentration of 1.7 ppm·m of CH4 in a path length of one meter in natural air, the system exhibits a high sensitivity for the detection of leaked methane.
From the concentration series, a calibration function is calculated for the measurement of path concentrations in an open-path geometry. Leakage is simulated with a removed gas cell by replacing the scattering target with a sand blasted aluminum plate. Through a 300-µm bore in the plate connected to the MFC, a gas flow can be induced. The MIR beam is aimed to impinge a few millimeters above the bore with the plated tilted slightly upwards such that incident beam and effused gas plume intersect. Figure 5(a) shows the configuration. Through the static plate, a gas flow of 1 % CH4 in N2 is switched on and off in a series with flow rates increasing from 3 to 60 ml/min. The simultaneously measured path concentrations are shown in Fig. 5(b). The stated path-concentration-values give the excess methane due to the leak in addition to the background methane in the lab air. The spatial distribution of the gas concentration is unknown, but the measurement shows that the integrated path concentration increases in correlation with the flow rate. Turbulences in the gas flow lead to fluctuating measured concentrations. However, a leakage CH4 signal can be well detected in the open path already at a flow rate of 3 ml/min, roughly corresponding to 1.3 µmol of methane effused in one minute.
Fig. 5 a) Sketch of the setup for a leak simulation with a punctuated aluminum target. b) Simulated static leak measurement. The induced flow rate and the measured path concentration are shown.
5. Conclusions
The presented system demonstrates efficient pulsed upconversion for the sensitive detection of mid-infrared signals. Synchronization of the MIR pulse generation with the pump pulses driving the upconversion is the main reason for the excellent system performance. Measured efficiency values show good agreement with theoretical predictions. The latter take into account the depletion of the pump pulse, which can lead to significant deviation from the simplified theoretical curve for high signal peak powers. Experimental data underlindes that pump-pulse depletion has to be considered at this stage.
Upconversion efficiencies of 82 % for collinear gaussian beam interaction and 25 % for the upconversion of a backscattered MIR spot over a sensitive numerical aperture of 0.05 are achieved. Due to the increased total upconversion efficiency compared to that of continuously driven conversion systems combined with the effective time gating and narrow spectral acceptance of the pulsed upconversion, as well as the low-noise detection on silicon devices, high signal-to-background ratios are obtained. This makes the approach very promising for sensitive measurements over large distances or on low-reflectivity targets, already at low average MIR signal powers, and especially implies robustness against elevated or fluctuating thermal background signals.
With the parametric generation of sub-nanosecond MIR pulses, a linewidth of 1.8 GHz, suitable for gas sensing, is achieved. The tunability of the pulsed source allows for precise spectroscopy of gas absorption lines. With measurements on a concentration series, good operating stability is demonstrated, leading to sensitive detection of CH4 path concentrations covering a high dynamic range down to below 1.5 ppm·m. The noise equivalent concentration is found to be 0.2 ppm·m·Hz−1/2. Compared to other reports of standoff gas sensing, as shown in Table 1, this states a low level of detection. Chosen examples are using a hard target standoff detection scheme. However, applied methods and conditions vary. With the achieved sensitivity, even small changes in the concentration, as would be caused for example by hidden or buried leaks, are well resolvable. Measurements undertaken with a simulated leak verify a detectable flow rate of just 3 ml/min of diluted methane.
Table 1. Standoff gas detection limits.
For a genuine open-path operation with variable target distance, a dynamic delay mechanism is required. Electronic synchronization of the MIR and pump pulse with actively Q-switched lasers could solve this issue [16] and even enable distance-scanning LIDAR applications, where the high detection bandwidth of near-infrared detectors enables a higher spatial resolution with sub-nanosecond light pulses.
Funding
Fraunhofer Internal Programs (MAVO 826 529); Fraunhofer and Max Planck cooperation programme.
References and links
1. BP, “Statistical world energy review,” (BP, 2016) http://www.bp.com/content/dam/bp/pdf/energy-economics/statistical-review-2016/bp-statistical-review-of-world-energy-2016-full-report.pdf
2. S. Solomon, Climate Change 2007 - The Physical Science Basis: Working Group I Contribution to the Fourth Assessment Report of the IPCC (Cambridge University, 2007).
3. B. van Well, S. Murray, J. Hodgkinson, R. Pride, R. Strzoda, G. Gibson, and M. Padgett, “An open-path, hand-held laser system for the detection of methane gas,” Journal of Optics A: Pure and Applied Optics 7(6), S420 (2005). [CrossRef]
4. G. B. Rieker, F. R. Giorgetta, W. C. Swann, J. Kofler, A. M. Zolot, L. C. Sinclair, E. Baumann, C. Cromer, G. Petron, and C. Sweeney, “Frequency-comb-based remote sensing of greenhouse gases over kilometer air paths,” Optica 1(5), 290–298 (2014). [CrossRef]
5. A. Fix, C. Büdenbender, M. Wirth, M. Quatrevalet, A. Amediek, C. Kiemle, and G. Ehret, “Optical parametric oscillators and amplifiers for airborne and spaceborne active remote sensing of CO2 and CH4,” Proc. SPIE 8182, 818206 (2011). [CrossRef]
6. B. H. Stuart, Infrared Spectroscopy: Fundamentals and Applications (Wiley, 2004).
7. A. Rogalski, Infrared Detectors (CRC Press, 2010).
8. D. Kleinman and G. Boyd, “Infrared detection by optical mixing,” J. Appl. Phys. 40(2), 546–566 (1969). [CrossRef]
9. P. Darré, R. Baudoin, J.-T. Gomes, N. Scott, L. Delage, L. Grossard, J. Sturmann, C. Farrington, F. Reynaud, and T. Ten Brummelaar, “First on-sky fringes with an up-conversion interferometer tested on a telescope array,” Phys. Rev. Lett. 117(23), 233902 (2016). [CrossRef] [PubMed]
10. Q. Hu, J. S. Dam, C. Pedersen, and P. Tidemand-Lichtenberg, “High-resolution mid-ir spectrometer based on frequency upconversion,” Optics Letters 37(24), 5232–5234 (2012). [CrossRef] [PubMed]
11. S. Wolf, J. Kiessling, M. Kunz, G. Popko, K. Buse, and F. Kühnemann, “Upconversion-enabled array spectrometer for the mid-infrared, featuring kilohertz spectra acquisition rates,” Opt. Express 25(13), 14504–14515 (2017). [CrossRef] [PubMed]
12. L. M. Kehlet, P. Tidemand-Lichtenberg, J. S. Dam, and C. Pedersen, “Infrared upconversion hyperspectral imaging,” Opt. Lett. 40(6), 938–941 (2015). [CrossRef] [PubMed]
13. J. S. Dam, P. Tidemand-Lichtenberg, and C. Pedersen, “Room-temperature mid-infrared single-photon spectral imaging,” Opt. Lett. 6(11), 788–793 (2012).
14. G. W. Faris and M. Banks, “Upconverting time gate for imaging through highly scattering media,” Opt. Lett. 19(22), 1813–1815 (1994). [CrossRef] [PubMed]
15. P. Geiser, U. Willer, D. Walter, and W. Schade, “A subnanosecond pulsed laser-source for mid-infrared lidar,” Appl. Phys. B 83, (2)175–179 (2006). [CrossRef]
16. L. Huot, P. M. Moselund, P. Tidemand-Lichtenberg, L. Leick, and C. Pedersen, “Upconversion imaging using an all-fiber supercontinuum source,” Opt. Lett. 41(11), 2466–2469 (2016). [CrossRef] [PubMed]
17. G. Millot, S. Pitois, M. Yan, T. Hovhannisyan, A. Bendahmane, T. W. Hänsch, and N. Picqué, “Frequency-agile dual-comb spectroscopy,” Opt. Lett. 10(1), 27–30 (2016).
18. R. P. Van Duyne, D. L. Jeanmaire, and D. Shriver, “Mode-locked laser Raman spectroscopy. New technique for the rejection of interfering background luminescence signals,” Anal. Chem. 46(2), 213–222 (1974). [CrossRef]
19. U. Bäder, T. Mattern, T. Bauer, J. Bartschke, M. Rahm, A. Borsutzky, and R. Wallenstein, “Pulsed nanosecond optical parametric generator based on periodically poled lithium niobate,” Opt. Commun. 217(1), 375–380 (2003). [CrossRef]
20. G. Marchev, F. Pirzio, R. Piccoli, A. Agnesi, G. Reali, P. G. Schunemann, K. T. Zawilski, A. Tyazhev, and V. Petrov, “Narrow-bandwidth, mid-infrared, seeded optical parametric generation in 90° phase-matched CdSiP2 crystal pumped by diffraction limited 500 ps pulses at 1064 nm,” Opt. Lett. 37(15), 3219–3221 (2012). [CrossRef] [PubMed]
21. M. M. Fejer, G. Magel, D. H. Jundt, and R. L. Byer, “Quasi-phase-matched second harmonic generation: Tuning and tolerances,” IEEE J. Quantum Electron. 28(11), 2631–2654 (1992). [CrossRef]
22. R. L. Sutherland, Handbook of Nonlinear Optics (CRC Press, 2003). [CrossRef]
23. A. Yariv, Quantum Electronics (Wiley, 1989).
24. L. Høgstedt, A. Fix, M. Wirth, C. Pedersen, and P. Tidemand-Lichtenberg, “Upconversion-based lidar measurements of atmospheric CO2,” Opt. Express 24(5), 5152–5161 (2016). [CrossRef]
25. L. S. Rothman, I. E. Gordon, Y. Babikov, A. Barbe, D. C. Benner, P. F. Bernath, M. Birk, L. Bizzocchi, V. Boudon, L. R. Brown, A. Campargue, K. Chance, E. A. Cohen, L. H. Coudert, V. M. Devi, B. J. Drouin, A. Fayt, J.-M. Flaud, R. R. Gamache, J. J. Harrison, J.-M. Hartmann, C. Hill, J. T. Hodges, D. Jacquemart, A. Jolly, J. Lamouroux, R. J. Le Roy, G. Li, D. A. Long, O. M. Lyulin, C. J. Mackie, S. T. Massie, S. Mikhailenko, H. S. P. Müller, O. V. Naumenko, A. V. Nikitin, J. Orphal, V. Perevalov, A. Perrin, E. R. Polovtseva, C. Richard, M. A. H. Smith, E. Starikova, K. Sung, S. Tashkun, J. Tennyson, G. C. Toon, V. G. Tyuterev, and G. Wagner, “The HITRAN2012 molecular spectroscopic database,” Journal of Quantitative Spectroscopy and Radiative Transfer 130, 4–50 (2013). [CrossRef]
26. M. Nikodem and G. Wysocki, “Chirped laser dispersion spectroscopy for remote open-path trace-gas sensing,” Sensors 12(12), 16466–16481 (2012). [CrossRef]
27. M. C. Phillips and B. E. Brumfield, “ECQCL developments for rapid standoff chemical sensing,” Proc. SPIE 10194, 101942T (2017). [CrossRef]
28. N. A. Macleod, R. Rose, and D. Weidmann, “Middle infrared active coherent laser spectrometer for standoff detection of chemicals,” Opt. Lett. 38(19), 3708–3711 (2013). [CrossRef] [PubMed]
