Cavity-enhanced optical frequency comb ... - OSA Publishing

Cavity-enhanced optical frequency comb spectroscopy of high-temperature H2O in a flame Chadi Abd Alrahman,1 Amir Khodabakhsh,1 Florian M. Schmidt,2 Zhechao Qu,2 and Aleksandra Foltynowicz1,* 2

1 Department of Physics, Umeå University, 901 87 Umeå, Sweden TEC-Lab, Department of Applied Physics and Electronics, Umeå University, 901 87 Umeå, Sweden *[email protected]

Abstract: We demonstrate near-infrared cavity-enhanced optical frequency comb spectroscopy of water in a premixed methane/air flat flame. The detection system is based on an Er:fiber femtosecond laser, a high finesse optical cavity containing the flame, and a fast-scanning Fourier transform spectrometer (FTS). High absorption sensitivity is obtained by the combination of a high-bandwidth two-point comb-cavity lock and autobalanced detection in the FTS. The system allows recording hightemperature water absorption spectra with a resolution of 1 GHz and a bandwidth of 50 nm in an acquisition time of 0.4 s, with absorption sensitivity of 4.2 × 10−9 cm−1 Hz-1/2 per spectral element. ©2014 Optical Society of America OCIS codes: (010.1030) Absorption; (300.6300) Spectroscopy, Fourier transforms; (120.6200) Spectrometers and spectroscopic instrumentation; (120.1740) Combustion diagnostics.

References and links 1.

J. J. Scherer and D. J. Rakestraw, “Cavity ringdown laser absorption spectroscopy detection of formyl (HCO) radical in a low pressure flame,” Chem. Phys. Lett. 265(1-2), 169–176 (1997). 2. J. Xie, B. A. Paldus, E. H. Wahl, J. Martin, T. G. Owano, C. H. Kruger, J. S. Harris, and R. N. Zare, “Nearinfrared cavity ringdown spectroscopy of water-vapor in an atmospheric flame,” Chem. Phys. Lett. 284(5-6), 387–395 (1998). 3. X. Zhou, J. B. Jeffries, and R. K. Hanson, “Development of a fast temperature sensor for combustion gases using a single tunable diode laser,” Appl. Phys. B 81(5), 711–722 (2005). 4. J. Vanderover and M. A. Oehlschlaeger, “A mid-infrared scanned-wavelength laser absorption sensor for carbon monoxide and temperature measurements from 900 to 4000 K,” Appl. Phys. B 99(1-2), 353–362 (2010). 5. R. S. Watt, T. Laurila, C. F. Kaminski, and J. Hult, “Cavity enhanced spectroscopy of high-temperature H2O in the near-infrared using a supercontinuum light source,” Appl. Spectrosc. 63(12), 1389–1395 (2009). 6. B. Lohden, S. Kuznetsova, K. Sengstock, V. M. Baev, A. Goldman, S. Cheskis, and B. Palsdottir, “Fiber laser intracavity absorption spectroscopy for in situ multicomponent gas analysis in the atmosphere and combustion environments,” Appl. Phys. B 102(2), 331–344 (2011). 7. F. Adler, M. J. Thorpe, K. C. Cossel, and J. Ye, “Cavity-enhanced direct frequency comb spectroscopy: Technology and applications,” Ann. Rev. Anal. Chem. (Palo Alto Calif) 3(1), 175–205 (2010). 8. R. Grilli, G. Mejean, S. Kassi, I. Ventrillard, C. Abd Alrahman, E. Fasci, and D. Romanini, “Trace measurement of BrO at the ppt level by a transportable mode-locked frequency-doubled cavity-enhanced spectrometer,” Appl. Phys. B 107(1), 205–212 (2012). 9. M. J. Thorpe, D. Balslev-Clausen, M. S. Kirchner, and J. Ye, “Cavity-enhanced optical frequency comb spectroscopy: application to human breath analysis,” Opt. Express 16(4), 2387–2397 (2008). 10. A. Foltynowicz, P. Maslowski, A. J. Fleisher, B. J. Bjork, and J. Ye, “Cavity-enhanced optical frequency comb spectroscopy in the mid-infrared - application to trace detection of hydrogen peroxide,” Appl. Phys. B 110(2), 163–175 (2013). 11. K. C. Cossel, F. Adler, K. A. Bertness, M. J. Thorpe, J. Feng, M. W. Raynor, and J. Ye, “Analysis of trace impurities in semiconductor gas via cavity-enhanced direct frequency comb spectroscopy,” Appl. Phys. B 100(4), 917–924 (2010).

#209012 - $15.00 USDReceived 27 Mar 2014; revised 22 May 2014; accepted 25 May 2014; published 30 May 2014 (C) 2014 OSA 2 June 2014 | Vol. 22, No. 11 | DOI:10.1364/OE.22.013889 | OPTICS EXPRESS 13889

12. S. Avino, A. Giorgini, M. Salza, M. Fabian, G. Gagliardi, and P. De Natale, “Evanescent-wave comb spectroscopy of liquids with strongly dispersive optical fiber cavities,” Appl. Phys. Lett. 102(20), 201116 (2013). 13. G. Hartung, J. Hult, and C. F. Kaminski, “A flat flame burner for the calibration of laser thermometry techniques,” Meas. Sci. Technol. 17(9), 2485–2493 (2006). 14. A. Foltynowicz, T. Ban, P. Masłowski, F. Adler, and J. Ye, “Quantum-noise-limited optical frequency comb spectroscopy,” Phys. Rev. Lett. 107(23), 233002 (2011). 15. L. S. Rothman, I. E. Gordon, A. Barbe, D. C. Benner, P. E. Bernath, M. Birk, V. Boudon, L. R. Brown, A. Campargue, J. P. Champion, K. Chance, L. H. Coudert, V. Dana, V. M. Devi, S. Fally, J. M. Flaud, R. R. Gamache, A. Goldman, D. Jacquemart, I. Kleiner, N. Lacome, W. J. Lafferty, J. Y. Mandin, S. T. Massie, S. N. Mikhailenko, C. E. Miller, N. Moazzen-Ahmadi, O. V. Naumenko, A. V. Nikitin, J. Orphal, V. I. Perevalov, A. Perrin, A. Predoi-Cross, C. P. Rinsland, M. Rotger, M. Simeckova, M. A. H. Smith, K. Sung, S. A. Tashkun, J. Tennyson, R. A. Toth, A. C. Vandaele, and J. Vander Auwera, “The HITRAN 2008 molecular spectroscopic database,” J. Quant. Spectrosc. Radiat. Transf. 110(9-10), 533–572 (2009). 16. L. S. Rothman, I. E. Gordon, R. J. Barber, H. Dothe, R. R. Gamache, A. Goldman, V. I. Perevalov, S. A. Tashkun, and J. Tennyson, “HITEMP, the high-temperature molecular spectroscopic database,” J. Quantum Spectrosc. Radiat. Transf. 111(15), 2139–2150 (2010).

1. Introduction Monitoring of parameters of combustion processes such as temperature and reactant/product concentrations is essential for the understanding of chemical reactions, enhancing combustion efficiency or reducing pollution levels. Laser-based absorption techniques are often used as a non-intrusive tool for combustion diagnostics, since they provide high sensitivity and high species selectivity in a short measurement time [1–4]. However, the limited tunability of continuous wave lasers, on which these techniques are based, implies that the concentration of only one or two species can be monitored with a particular system. It also often imposes restrictions on the choice of absorption line pairs that can be used for flame thermometry [3]. Broadband absorption measurements have been performed in flames using cavity-enhanced absorption spectroscopy with a supercontinuum (SC) radiation source [5] and fiber laser intracavity absorption spectroscopy (FLICAS) [6]. The SC approach required relatively long integration times (on the order of tens of minutes) to achieve high signal-to-noise ratio. In FLICAS the simultaneous spectral coverage did not exceed 10 nm and tuning of the laser parameters was required to cover wider spectral ranges. Moreover, both techniques provided a rather low spectral resolution (a few GHz, limited by the spectrometers) and required including the instrument function in models of absorption lines. Cavity-enhanced optical frequency comb spectroscopy (CE-OFCS) combines the broad spectral bandwidth and high spectral resolution of an optical frequency comb (OFC) with high detection sensitivity provided by the use of enhancement cavities and efficient noise reduction schemes [7]. The potential of CE-OFCS has already been demonstrated in the gas phase for detection of molecular species important for environmental research [8], human breath analysis [9, 10], and monitoring of production processes [11], as well as in the liquid phase [12]. Here we employ CE-OFCS for the first time for measurements of near-infrared water absorption spectra in a laminar, premixed methane/air flat flame. The system is capable of acquiring tens of nm of high resolution spectra with high signal-to-noise ratio in times below a second. High power throughput and stable cavity transmission are obtained by tightly locking an Er:fiber comb to the cavity that contains the flame. Absorption spectra are recorded using a fast-scanning Fourier transform spectrometer with a noise reduction scheme based on an auto-balancing detector. The stable comb-cavity lock allows uninterrupted measurements at different heights above the burner and opens up for real-time characterization of flame zones. 2. Experimental setup The experimental setup, illustrated in Fig. 1, is based on an Er:fiber femtosecond laser, a high finesse optical cavity containing the flame, and a fast-scanning Fourier transform spectrometer (FTS). For technical reasons, the cavity and the mode-matching optics are placed in a different room than the rest of the setup, and the two parts of the system are connected by 8 m long polarization maintaining (PM) fibers. The Er:fiber laser provides an


average power of 20 mW in the 1.5-1.6 µm range with a repetition rate, fr, of 250 MHz. The laser fr is controlled by a piezoelectric transducer (PZT) and a stepper motor that change the laser cavity length, and an intracavity electro-optic modulator (EOM) that modulates the intracavity refractive index. The carrier-envelope offset frequency, f0, of the comb is controlled via the current of the pump diode laser. The cavity is made of two concave mirrors (M, with radius of curvature 5 m) with specified reflectivity of 99.7% at 1530 nm. The cavity length is 60 cm to match the free spectral range to the laser fr. Because of the relatively weak absorption of atmospheric H2O and CO2 in the operating wavelength range of the laser, the cavity can be open to air and purging of the laser beam path is not necessary. The burner, based on the design described in [13], Hartung et al, is mounted on a vertical translation stage below the cavity so that the flame is centered between the cavity mirrors. This configuration allows measurements of spectra at different heights above the burner (HAB). The diameter of the flame is 3.8 cm and the intracavity beam radius at the position of the flame is 0.77 mm. The premixed CH4/air flame is operated at a stoichiometric ratio of 1 with flow rates of 1 and 10 l/min for CH4 and air, respectively. Under these conditions, the combustion zone is about 1 mm above the burner plate, and homogeneous, close to adiabatic, conditions can be expected at HAB of 5 mm and above [13]. An annular N2 co-flow is supplied through a concentric ring in the burner to shield the flame from the ambient environment and thus stabilize it.

Fig. 1. Experimental setup. EOM: electro-optic modulator, PMF: polarization maintaining fiber, FC: fiber collimator, HWP: half-wave plate, PBS: polarizing beam splitter cube, f: lens, QWP: quarter-wave plate, M: cavity mirror, PDH lock: error signal detection, FTS: Fourier transform spectrometer.

The output of the comb is phase modulated at 20 MHz by a PM-fiber-coupled EOM to generate sidebands for locking of the comb to the cavity. The modulated light is led via a PM fiber (PMF1) to the cavity setup, coupled out into free space, and spatially mode-matched to the cavity TEM00 mode with two lenses (f). The beam reflected from the cavity is picked off with a combination of a polarizing beam splitter (PBS) cube and a quarter-wave plate (QWP) and coupled into another PM fiber (PMF2). A half-wave plate (HWP) in front of the PBS is used to adjust the polarization for maximum transmission through the PBS. The resulting optical power incident on the cavity is 6.5 mW. The comb is locked to the cavity by the two-point locking scheme [10], in which error signals are derived from the cavity reflected light at two different wavelengths (referred to as locking points) using the Pound-Drever-Hall (PDH) technique. Each PDH error signal is used to lock a group of comb lines to their respective resonant cavity modes. The error signal from the first locking point is pre-amplified by a proportional-integral (PI) servo and fed in parallel to two separate PI servos controlling the laser PZT and EOM. The error signal from the second locking point is fed to another PI servo that controls the pump diode current. The closed-loop bandwidths are 500 kHz for the PZT + EOM lock and 200 kHz for the current lock. When the flame is on, an additional slow and large range feedback is sent to a stepper motor (not shown in the figure) in order to account for the thermal drift of the cavity length, which causes a change of fr that is larger than what can be compensated by the laser PZT. It is worth to note that we achieve sufficiently high locking bandwidth even with the use of an 8 m PM fiber that connects the cavity and the error signal detection unit.


The beam transmitted through the cavity is mode-matched into a PM fiber (PMF3) using a pair of lenses, and injected into the FTS. A QWP is used after the cavity to create a linear polarization state aligned to the direction of the key of the PMF3. The design of the FTS is similar to that described in [14], Foltynowicz et al. The outputs of the interferometer are monitored with an auto-balancing detector based on InGaAs photodiodes, whose output is digitized by a high-resolution data acquisition card. To avoid saturation, the optical power is attenuated by a factor of ~3 in front of the FTS so that the power incident on each photodiode is ~300 µW. The optical path difference is calibrated using a stabilized HeNe laser (with stability on the order of 10−7), whose beam is propagating parallel to the OFC beam in the FTS. Fast Fourier transform of the comb interferogram, resampled at the zero crossings and extrema of the HeNe interferogram, yields the broadband cavity transmitted spectrum. Since all measurements are performed at atmospheric pressure, where absorption lines have pressure broadened linewidths on the order of a few GHz, the resolution is set to 1 GHz. The high scanning speed of the FTS (0.2 m/s) allows acquisition of one spectrum in 0.4 s. All spectra in the results section are interpolated to 250 MHz by zero padding of the interferogram. 3. Results and discussion In order to couple a large fraction of the comb spectrum into the cavity we set the locking points at 1550 nm for the current lock and 1575 nm for the PZT + EOM lock. The resulting background spectrum (i.e. when the flame is off) is shown by the black curve in Fig. 2(a). In general, 50 nm of spectrum centered at 1560 nm is transmitted and the number of resolved spectral elements is ~6000; the transmitted bandwidth is limited by dispersion in cavity mirror coatings. The red curve in Fig. 2(a) shows a spectrum recorded at HAB of 5 mm when the flame is on. A dense spectrum of high-temperature water lines is clearly visible. The overall decrease of intensity is caused partly by water absorption and partly by beam steering effects inside the flame. A zoom of ten consecutive cavity background spectra (black) and ten hightemperature water spectra recorded at HAB of 1 mm (blue) and 5 mm (red) is shown in Fig. 2(b). The weak absorption lines in the background spectrum are from atmospheric CO2 at room temperature. The stability of the spectra recorded in the flame is slightly worse than that of the background spectra because of the pointing instability of the intracavity beam, but in general the reproducibility of the spectra is very good.

Fig. 2. (a) Background spectrum transmitted through the cavity at room temperature, i.e. when the flame is off (black), and high-temperature H2O absorption spectrum with the flame on measured at HAB of 5 mm (red). Arrows indicate the locking points. (b) Zoom of ten consecutive cavity background spectra (black), and ten consecutive high-temperature H2O absorption spectra measured at HAB of 1 mm (blue) and 5 mm (red).

We estimate the noise equivalent absorption (NEA) by taking the ratio of two consecutive background spectra, fitting and removing a 3rd order polynomial baseline, and calculating the standard deviation of the noise on the resulting baseline in the middle of the transmitted spectrum. The relative noise is 9.6 × 10−4, which yields NEA in a single element of the spectrum of 2.3 × 10−8 cm−1 in 0.8 s [calculated as the ratio of the noise and the effective


interaction length inside the cavity, given by 2FL/π, where F = 1100 is the cavity finesse at 1560 nm, see Fig. 3(a), and L is the cavity length]. The NEA per spectral element is thus 2.6 × 10−10 cm−1 Hz-1/2 for a spectrum recorded at room temperature. Since the flame diameter is 3.8 cm, the NEA for the high-temperature spectra is 4.2 × 10−9 cm−1 Hz-1/2 per spectral element. To demonstrate the performance of the system we compare the absorption spectra of room temperature atmospheric H2O and CO2, visible in the background spectrum, to a theoretical model. We calculate the theoretical spectra using absorption line parameters from the HITRAN database [15], a Voigt profile for the individual absorption lines, and a model of cavity-enhanced absorption given by Eq. (4) in [10], Foltynowicz et al. Since the spectral range of interest lies between the two locking points, we set the frequency detuning of comb lines from the center of the corresponding cavity modes to zero [10]. We also take into account the wavelength dependence of the cavity finesse, which we measured by cavity ringdown. The experimentally obtained finesse values are shown by blue markers in Fig. 3(a), while the red curve shows a 3rd order polynomial fit to the data that we use in the model of the spectrum. In order to normalize the spectrum, we reconstruct the baseline by a sum of sine functions whose frequencies are found via a sine transform of the background spectrum. Finally, we fit the sum of the models of absorption spectra and the baseline to the experimental spectrum, with concentrations of the two gases and the amplitudes of the sine functions as fitting parameters. Figure 3(b) shows the absorption spectrum in air (black, 30 averages) normalized to the fitted baseline, and the fitted H2O (blue) and CO2 (red) spectra, inverted and offset for clarity. The concentrations of H2O and CO2 are found to be 0.48% and 621 ppm, respectively. A comparison of the theoretical model to the experimental data shows that the lineshapes of CO2 and H2O are reproduced well over the entire spectral range. The remaining structure in the residual is caused by inaccuracies in the HITRAN database (for the weak H2O lines), and by the neglected small comb-cavity mode detuning (the dispersive residual for CO2 lines).

Fig. 3. (a) Cavity finesse (blue markers) with a 3rd order polynomial fit (red). (b) Normalized absorption spectrum of atmospheric H2O and CO2 at room temperature (black), and fitted theoretical spectra of H2O (blue) and CO2 (red), inverted and offset for clarity. Bottom panel shows the residual of the fit.

Figure 4(a) shows the high-temperature H2O spectrum recorded in the flame at HAB of 5 mm normalized to the cavity background spectrum taken before turning the flame on (black, 1 average), i.e. the ratio of the two spectra shown in Fig. 2(a). The red curve (inverted for clarity) shows a cavity-enhanced H2O spectrum calculated using the same model as above and the absorption line parameters from the HITEMP database [16] for a temperature of 2000 K and a water concentration of 18% (conditions expected at this HAB [13]). The absorption of CO2 at this temperature is much lower than that of H2O and can be neglected. In order to match the model to the experimental data we multiplied the normalized experimental spectrum by a factor of 1.2 to compensate for the intensity decrease caused by the beam pointing instability. After this correction the general agreement between the model and the experimental data is good. However, a close inspection reveals a mismatch in line positions and intensities, as shown in a zoom of a spectral window around 1549 nm in Fig. 4(b). We


attribute these discrepancies to inaccuracies in the line parameters listed in the HITEMP database rather than to any systematic errors in the measurement. Many lines within the acquired spectral range exhibit strong temperature dependence. As an example, Fig. 4(c) shows a zoom of normalized (and multiplied by 1.2) H2O spectra around 1572 nm recorded at HAB between 1 mm and 5 mm (30 averages each). This HAB interval covers the transition between the combustion and the post-combustion zones, where an increase in temperature and H2O concentration is expected. The increase in concentration is reflected in the overall increase of absorption with HAB, while strong temperature dependence is visible in two groups of lines (at 1572.14 and 1572.4 nm). The lack of an accurate model of high-temperature water spectra in this wavelength range prevents at this stage quantitative analysis of water concentration and flame temperature. However, the stability and robustness of the system (measurements at different HAB are made without unlocking the comb from the cavity) will allow systematic measurements whose results can be compared to those obtained with direct absorption spectroscopy and simulations, provided a methodology to discriminate between the pointing instability and broadband absorption is implemented. This will be the subject of a future study.

Fig. 4. (a) Normalized high-temperature H2O absorption spectrum at HAB of 5 mm in a flame (black) with a model based on parameters from the HITEMP database (red) for temperature of 2000 K and water concentration of 18%, inverted for clarity. (b) A zoom of (a) around 1549 nm, where the theoretical spectrum is not inverted. (c) A zoom around 1572 nm of H2O absorption spectra taken at different HAB.

4. Conclusions We have demonstrated that CE-OFCS with a tight comb-cavity lock can be employed for measurements of broadband high-temperature spectra in a combustion environment with high signal-to-noise ratio in acquisition times below a second. The use of a cavity allows addressing spectral ranges with weak absorption lines, where the interference from atmospheric species is negligible. The high bandwidth and large control range of the lock ensure uninterrupted operation even as the conditions in the flame change. The use of PM fibers to connect the cavity setup to the comb source, error signal detection unit, and the FTS does not deteriorate the performance; on the contrary, it offers the advantage of performing measurements at remote locations. The resolution of the spectra is limited by the pressure broadening of the absorption lines, and the spectrometer does not introduce any distortion to the experimental spectra. If needed, e.g. for measurements at lower pressures, the resolution can be improved by orders of magnitude by resolving the comb lines. The successful realization of CE-OFCS in a flame opens up the possibility of real-time multispecies detection in a variety of combustion environments. Extending the use of the technique to other wavelength ranges (especially to the mid-infrared) via nonlinear frequency conversion will enable simultaneous detection of a long list of molecules, e.g. hydrocarbons in biomass combustion processes. The technique can also allow determination of flame temperature with high accuracy by the use of an entire absorption spectrum instead of a pair of absorption lines. #209012 - $15.00 USDReceived 27 Mar 2014; revised 22 May 2014; accepted 25 May 2014; published 30 May 2014 (C) 2014 OSA 2 June 2014 | Vol. 22, No. 11 | DOI:10.1364/OE.22.013889 | OPTICS EXPRESS 13894

Acknowledgments This project was supported by the Swedish Research Council (621-2012-3650), Swedish Foundation for Strategic Research (ICA12-0031), Swedish Energy Agency (36160-1), Umeå University, Carl Tryggers Stiftelse (12:131), and Stiftelsen Lars Hiertas Minne. The authors thank Grzegorz Kowzan for implementing the software for repetition rate control.
