Intra-pulse laser absorption sensor with cavity enhancement for oxidation experiments in a rapid compression machine

A sensor based on a mid-IR pulsed quantum cascade laser (QCL) and off-axis cavity enhanced absorption spectroscopy (OA-CEAS) has been developed for highly sensitive concentration measurements of carbon monoxide (CO) in a rapid compression machine. The duty cycle and the pulse repetition rate of the laser were optimized for increased tuning range, high chirp rate, and small line width to achieve effective laser-cavity coupling. This enabled spectrally resolved CO line-shape measurements at high pressures (P ~10 bar). A gain factor of 133 and a time resolution of 10 μs were demonstrated. CO concentration-time profiles during the oxidation of highly dilute n-octane/air mixtures were recorded, illustrating new opportunities in RCM experiments for chemical kinetics. © 2018 Optical Society of America under the terms of the OSA Open Access Publishing Agreement OCIS codes: (280.1740) Combustion diagnostics; (280.3420) Laser sensors; (300.1030) Absorption. References and links 1. R. K. Hanson, “Applications of quantitative laser sensors to kinetics, propulsion and practical energy systems,” Proc. Combust. Inst. 33(1), 1–40 (2011). 2. C. J. Sung and H. J. Curran, “Using rapid compression machines for chemical kinetics studies,” Pror. Energy Combust. Sci. 44, 1–18 (2014). 3. K. Sun, S. Wang, R. Sur, X. Chao, J. B. Jeffries, and R. K. Hanson, “Sensitive and rapid laser diagnostic for shock tube kinetics studies using cavity-enhanced absorption spectroscopy,” Opt. Express 22(8), 9291–9300 (2014). 4. D. Romanini, I. Ventrillard, G. Méjean, J. Morville, and E. Kerstel, “Introduction to Cavity Enhanced Absorption Spectroscopy,” in Cavity-Enhanced Spectroscopy and Sensing, G. Gagliardi and H.-P. Loock, eds. (Springer Berlin Heidelberg, Berlin, Heidelberg, 2014), pp. 1–60. 5. S. Wang, K. Sun, D. F. Davidson, J. B. Jeffries, and R. K. Hanson, “Cavity-enhanced absorption spectroscopy with a ps-pulsed UV laser for sensitive, high-speed measurements in a shock tube,” Opt. Express 24(1), 308–318 (2016). 6. K. Sun, S. Wang, R. Sur, X. Chao, J. B. Jeffries, and R. K. Hanson, “Time-resolved in situ detection of CO in a shock tube using cavity-enhanced absorption spectroscopy with a quantum-cascade laser near 4.6 μm,” Opt. Express 22(20), 24559–24565 (2014). 7. G. Maisons, P. G. Carbajo, M. Carras, and D. Romanini, “Optical-feedback cavity-enhanced absorption spectroscopy with a quantum cascade laser,” Opt. Lett. 35(21), 3607–3609 (2010). 8. E. F. Nasir and A. Farooq, “Time-resolved temperature measurements in a rapid compression machine using quantum cascade laser absorption in the intrapulse mode,” Proc. Combust. Inst. 36(3), 4453–4460 (2017). 9. D. S. Sayres, E. J. Moyer, T. F. Hanisco, J. M. St. Clair, F. N. Keutsch, A. O’Brien, N. T. Allen, L. Lapson, J. N. Demusz, M. Rivero, T. Martin, M. Greenberg, C. Tuozzolo, G. S. Engel, J. H. Kroll, J. B. Paul, and J. G. Anderson, “A new cavity based absorption instrument for detection of water isotopologues in the upper troposphere and lower stratosphere,” Rev. Sci. Instrum. 80(4), 044102 (2009). 10. G. S. Engel, W. S. Drisdell, F. N. Keutsch, E. J. Moyer, and J. G. Anderson, “Ultrasensitive near-infrared integrated cavity output spectroscopy technique for detection of CO at 1.57 microm: new sensitivity limits for absorption measurements in passive optical cavities,” Appl. Opt. 45(36), 9221–9229 (2006). 11. R. S. M. Chrystie, E. F. Nasir, and A. Farooq, “Ultra-fast and calibration-free temperature sensing in the intrapulse mode,” Opt. Lett. 39(23), 6620–6623 (2014). 12. M. T. McCulloch, E. L. Normand, N. Langford, G. Duxbury, and D. A. Newnham, “Highly sensitive detection of trace gases using the time-resolved frequency downchirp from pulsed quantum-cascade lasers,” J. Opt. Soc. Am. B 20(8), 1761–1768 (2003). 13. A. B. S. Alquaity, U. Kc, A. Popov, and A. Farooq, “Detection of shock-heated hydrogen peroxide (H2O2) by off-axis cavity-enhanced absorption spectroscopy (OA-CEAS),” Appl. Phys. B 123(12), 280 (2017). Vol. 26, No. 11 | 28 May 2018 | OPTICS EXPRESS 14601 #327971 https://doi.org/10.1364/OE.26.014601 Journal © 2018 Received 11 Apr 2018; revised 18 May 2018; accepted 19 May 2018; published 23 May 2018 14. 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, J. Tennyson, G. C. Toon, H. Tran, V. G. Tyuterev, A. Barbe, A. G. Császár, V. M. Devi, T. Furtenbacher, J. J. Harrison, J. M. Hartmann, 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. V. Auwera, G. Wagner, J. Wilzewski, P. Wcisło, S. Yu, and E. J. Zak, “The HITRAN2016 molecular spectroscopic database,” J. Quant. Spectrosc. Radiat. Transf. 203, 3–69 (2017). 15. U. Kc, E. F. Nasir, and A. Farooq, “A mid-infrared absorption diagnostic for acetylene detection,” Appl. Phys. B 120(2), 223–232 (2015). 16. R. M. Spearrin, C. S. Goldenstein, J. B. Jeffries, and R. K. Hanson, “Quantum cascade laser absorption sensor for carbon monoxide in high-pressure gases using wavelength modulation spectroscopy,” Appl. Opt. 53(9), 1938–1946 (2014). 17. S. M. Sarathy, C. K. Westbrook, M. Mehl, W. J. Pitz, C. Togbe, P. Dagaut, H. Wang, M. A. Oehlschlaeger, U. Niemann, K. Seshadri, P. S. Veloo, C. Ji, F. N. Egolfopoulos, and T. Lu, “Comprehensive chemical kinetic modeling of the oxidation of 2-methylalkanes from C7 to C20,” Combust. Flame 158(12), 2338–2357 (2011).


Introduction
Experimental investigations of high-temperature (T > 1000 K) gas-phase kinetics have been substantially augmented by measurement techniques based on laser absorption spectroscopy [1].Such experiments are usually performed under controlled thermodynamic conditions in homogeneous environments, such as shock tubes and rapid compression machines (RCM).These conditions are perfectly suited to path-averaged laser absorption techniques which can achieve high sensitivity in species concentration measurements by means of strong rovibrational transitions in the mid-infrared wavelength region.Recent advances in combustion research have shifted attention to lower temperature kinetics (700 K < T < 900 K) in RCMs [2], wherein slower chemical processes result in much lower species production than at higher temperatures, thus putting greater sensitivity demands on measurement methods.For a given set of thermodynamic conditions and a particular molecular absorption transition, the Beer-Lambert law for absorption spectroscopy provides only one option for maximizing the measured laser absorption signal, which is to increase the beam path length within the probe gas volume [3].Optically resonant cavities are the most efficient means of achieving manifold increase in effective path length and, hence, the sensitivity of an existing laser absorption diagnostic [4].
In this work, we present an implementation of a low-finesse optical cavity in an RCM for time-resolved measurements of carbon monoxide concentration via cavity enhanced absorption spectroscopy (CEAS).We also present the use of spectral frequency down-chirp phenomenon in pulsed quantum cascade lasers (QCL) for minimizing laser-cavity coupling noise.Effective cavity coupling may be achieved via several methods: using a broadband laser source such that the FWHM of the laser mode covers many cavity modes, as demonstrated with a ps-pulsed Ti-sapphire UV laser by Wang et al. [5]; rapid scanning of a narrow linewidth laser such that the laser mode traverses several cavity modes within the time resolution of the detection system, as shown by Sun et al. [6] with a cw-QCL; and active locking of the cavity through optical feedback [7].In our case, pulsed mode operation of the QCL affords us both a broader line-width than cw-QCLs as well as a rapid scan capability to resolve molecular line shapes.Our implementation also avoids the experimental complexity of active mode-locking via optical feedback to an RCM-mounted cavity by utilizing an offaxis beam alignment, which reduces the cavity free spectral range (FSR) and enhances spurious mode coupling.

Cavity characteristics
The optical cavity was set up on the twin-opposed piston RCM facility (described previously in [8]) using two concave mirrors (supplied by Rocky Mountain Instruments), as described in Fig. 1.The RCM combustion chamber has an inner diameter of 5.08 cm, which serves as the single-pass path length (L) in this case.The mirrors had a 100 cm radius of curvature (R m ) and a specified nominal reflectivity of 99.5% ± 0.2% at 4.887 μm, which is the center wavelength (λ) of the QCL used in this work (also described in [8]).Reflective coating on the mirrors also provided similar reflectivity at 633 nm to aid alignment with a visible laser.The actual mirror reflectivity (R) was experimentally found to be 99.25% ± 0.04% (see section 3.1 of this article).Based on the path-length, reflectivity and a characteristic single-pass absorbance (A SP ) of 0.01, the characteristic photon residence time (τ) for the cavity is calculated to be 9.67 ns via the following relation: (1 ) The QCL was aligned in an off-axis arrangement to create a Lissajous spot pattern on the cavity mirrors, in a manner described by Sayres et al. [9] and illustrated in Fig. 1.The laser signal was recorded on a high bandwidth AC-coupled HgCdTe detector (from Vigo System).An aspheric lens with small focal length (12.7 mm) was used to collect the beam pattern exiting the RCM cavity and focus onto the 1 mm 2 active area of the detector.The nominal spot size (s) for this cavity is 1.2 mm, calculated using the following relation [9].
The diameter of the mirrors used in this work was 12.7 mm; however, due to the nature of the mirror mounting procedure, which involved seating the mirrors on the cavity plugs with epoxy, the actual available optical area on the mirrors is about 8 mm.During laser alignment, the spot pattern was limited to a circular area of 5 mm diameter to avoid it being clipped from the edge of the optical plug in the event of significant mechanical vibration.Assuming a circular spot pattern, the maximum number of non-overlapping spots that can fit within a 5 mm diameter area is 13.The effective free spectral range (FSReff) of the cavity can be calculated as [9] (assuming a refractive index of unity): Here, c is the speed of light and n refers to the number of spots which are equal to the number of passes by the laser beam before laser re-entry.For an on-axis alignment, n = 1, and in this case, the FSR is 2964 MHz.For the off-axis arrangement described above, n = 13; therefore, the effective cavity FSR is reduced to 227 MHz.A reduced cavity FSR increases the density of the cavity mode structure within the same spectral frequency window and leads to lower cavity coupling noise than an on-axis arrangement [10].

Pulsed-QCL characteristics
The QCL was used to probe the P(23) ro-vibrational transition in the fundamental band of CO.For pulsed mode operation, it is necessary to select the optimal pulse duty cycle (ratio of pulse duration to repetition period) and repetition rate.Three considerations need to be made: 1) sufficient measurement time resolution for chemical processes in the RCM; 2) spectrally resolved CO absorption line-shapes at high pressures (P ~10 bar); 3) cavity coupling noise suppression via chirp-induced laser line broadening.Chrystie et al. [11] had previously shown that for pulsed QCLs, the tuning range decreases with increasing repetition rate, and a maximum tuning range is obtained at 50% duty cycle for a specified repetition rate.A similar characterization of tuning range was performed for the QCL used in this work.The laser beam was split using a beam-splitter with one part passing through a Germanium etalon and the other through the CEAS cavity in an on-axis alignment.Figure 2(a) shows intensity traces for one case at 100 kHz repetition rate and 500 ns pulse duration (5% duty cycle).The tuning range across the pulse is determined from the fringes produced by the Germanium etalon.The measured on-axis cavity FSR, determined from the cavity fringes and the previously determined tuning range, is consistent with the calculated value of 2964 MHz.For pulse repetition rates between 100 kHz -1.5 MHz, the measured tuning range follows the trends described by Chrystie et al. [11], as illustrated in Fig. 2(b).Chemical processes in RCM experiments generally occur on the millisecond time-scale; therefore, a repetition rate of 100 kHz would result in a sufficiently high measurement time resolution (10 μs) while also providing a higher tuning range.Although a 50% duty cycle may seem to be the optimal operating point from Fig. 2(b), consideration of laser-cavity coupling noise yields a different outcome.Figure 3 shows measured off-axis laser intensity traces for duty cycles ranging from 1 -50%, which correspond to 100 -5000 ns pulse duration for 100 kHz repetition rate.As pulse duration increases, the cavity coupling noise is also found to increase, becoming more pronounced towards the latter part of the pulse.This would suggest that frequency down-chirp, a key determinant of cavity noise suppression, is varying within the pulse.Using the previously recorded Germanium etalon traces, the change in chirp rate with time was determined through the relative fringe spacing.The resulting laser line-width (Δν) for QCLs is related to the chirp rate (dν/dt) via the following expression [12]: The constant β has an upper limit value of unity, which is used here to estimate the laser linewidth.Figure 4 shows the variation of (a) chirp rate and (b) laser line-width within the laser pulse for different pulse durations at 100 kHz.It is seen in Fig. 3 that the coupling noise is effectively suppressed for pulses up to 500 ns, whereas a build-up in coupling noise is observed for 1000 ns pulse and the noise reaches an unacceptable level for 5000 ns pulse.The noise build-up may be explained by the decrease in chirp rate, and hence laser line-width, across the pulse, as illustrated in Fig. 4. At 1000 ns, the laser line-width decreases to the value of the off-axis cavity FSR (227 MHz).Given that this is approximately when the coupling noise build-up starts, the off-axis cavity FSR serves as a noise-limiting lower limit for the QCL line-width.Based on the consideration of signal-tonoise ratio, 500 ns was selected as the optimal pulse duration for the RCM experiments.The average laser line-width obtained at this operating point is 389 MHz (0.013 cm −1 ) which is significantly higher than the off-axis cavity FSR (shown in Fig. 5(a)) and is negligible compared to the FWHM (~1 cm −1 ) of the CO absorption line at 10 bar.Moreover, the average chirp rate of 154 MHz/ns for 500 ns pulse duration causes the laser emission mode to shift by 1485 MHz within the photon residence time of the cavity, as illustrated in Fig. 5(b).Therefore, on average, the laser mode traverses one cavity mode every 1.5 ns.A low-pass signal filter, with a larger rise time than 1.5 ns, would result in a time-averaged smoothing of the acquired laser signal [13].In our case, we applied a 20 MHz low-pass filter, resulting in a rise time of 17.5 ns which is significantly smaller than the pulse duration (500 ns) and yet large enough to have the aforementioned smoothing effect.Finally, the chosen 500 ns pulse duration also fulfills the consideration of covering the CO absorption line at 10 bar with a sufficiently large tuning range (3.26 cm −1 ).

Gain determination
For the CO mole fraction to be determined from the measured fractional laser intensity transmission, the CEAS absorbance A CEAS , equal to ln(I 0 /I t ), must first be converted to an equivalent single-pass absorbance A SP via the following relation [13]: The gain factor (G) is a property of the cavity and is derived from the mirror reflectivity through the following: The manufacturer-specified values for mirror reflectivity, however, cannot be used directly for calculating G because the uncertainty bounds on G (143 -333), computed from the specified uncertainty on mirror reflectivity ( ± 0.2%), are unacceptably large.Secondly, the nominal value of mirror reflectivity can deviate significantly from the manufacturer's specification, as reported previously by Alquaity et al. [13].A precise value of G must therefore be determined experimentally.We performed calibration experiments for G using mixtures of CO/N 2 with known compositions.The gain factor was found by using simulated peak A SP from HITRAN line parameters [14] in Eq. ( 5) along with measured peak A CEAS .The final value for gain based on several experiments was determined to be 133 ± 8, with significant reduction of the uncertainty bounds determined from the manufacturer-specified uncertainty in mirror reflectivity.The measured gain translates to a mirror reflectivity of 99.25 ± 0.04%.Figure 6(a) shows measured laser intensity traces for a representative gain factor calibration experiment conducted at 7 bar and 353 K, from which the single-pass absorbance spectrum is produced in Fig. 6(b) where the line-shape is fit with a Voigt profile.This experiment was performed within the RCM but under static conditions, i.e., the RCM combustion chamber was filled with high pressure gas instead of gas being compressed by the pistons.Since these were static steady-state conditions, the data were averaged over 512 readings, yielding a smoother signal.The line-strength determined from the integrated area of the Voigt fit was found to be in close agreement with HITRAN value (within 2%).It should be noted that the gain was only determined using high pressure gas mixtures (P > 5 bar) due to the need for the laser line-width to be significantly smaller than the CO collisional width.For experiments where CO concentration was to be measured, the concentration may be determined from either the integrated area or the peak absorbance.We found that the use of peak absorbance resulted in a much higher signal to noise ratio in the measured CO timehistories, consistent with a previous conclusion by KC et al. [15].Subsequently, experimental CO concentration profiles were obtained using the peak ASP.The sensor was then tested in dynamic tests in the RCM, i.e., with piston compression, at a nominal pressure of 10 bar.The gain was monitored after each subsequent reactive experiment with a CO/N 2 mixture to determine if any reflectivity degradation occurs or if changes in the wall temperature of the RCM has any effect.In all of these cases, the gain was found to be consistent (within uncertainty limits) with the value determined from static tests.
The minimum detection limit can be determined based on the minimum detectable absorbance.For a minimum detectable absorbance of 0.01, the minimum detection limit of a single-pass CO measurement using the P(23) CO line is 151 ppm at 800 K and 10 bar.In a CEAS experiment, however, the measured absorbance is converted to an equivalent singlepass absorbance.Thus a CEAS absorbance of 0.01 will yield a much lower single-pass absorbance and hence a much lower detection limit.For 800 K and 10 bar, this results in a minimum CO detection limit of 2.4 ppm based on a gain factor of 133.The CEAS technique, thus, clearly provides a significant improvement in sensitivity over conventional direct absorption.The CEAS method also provides significantly greater sensitivity than wavelength modulation spectroscopy, which improves SNR by filtering low-frequency noise.In a previous study by Spearrin et al. [16], a different P-branch CO line was used to obtain a minimum detection limit of ~500 ppm at similar conditions (L = 4 cm, P = 10 bar, T = 1200 K).While the WMS technique may improve a direct absorption measurement by filtering out sources of low-frequency noise such as window fouling and mechanical vibrations, the CEAS method delivers orders of magnitude improvement in mole fraction sensitivity than a direct absorption experiment.

n-Octane oxidation experiments
The laser diagnostic was used to measure CO mole fraction time-history during reactive experiments in the RCM for stoichiometric (φ = 1) n-octane/air mixtures.Conditions in the RCM are typically tailored to be approximately adiabatic and homogeneous [2], wherein lineof-sight absorption measurements can be applicable.A non-homogeneous temperature field can arise from substantial heat release during oxidation experiments; therefore, the n-octane mole fraction in the gas mixture was kept relatively low (0.2%).A reduced fuel concentration ensured that heat release during oxidation would cause a negligible temperature rise.The chosen mixture would only undergo partial oxidation (first-stage ignition) due to the diluted conditions but would still result in CO formation, thereby yielding important information for low-temperature reactions.
Figure 7(a) shows measured pressure (black) and peak CEAS absorbance (ACEAS) timehistory (blue) for a reactive experiment in the RCM.It should be noted that this is a transient, single-shot measurement.The pressure rise occurs due to the adiabatic compression of the gas mixture, which also raises the temperature.The maximum point of pressure after compression is designated "time zero", beyond which, a high temperature/pressure environment exists in the RCM combustion chamber.The temperature time-history is inferred from the measured pressure trace using the isentropic gas relations [8].The high-frequency noise in the peak ACEAS profile is mainly due to the pulse-to-pulse intensity variation and was removed using a low-pass 2 kHz filter, whereas the low-frequency noise is due to mechanical vibrations within the RCM affecting the mirror plugs.The smoothed A CEAS time-history was converted to CO mole fraction time-history shown in Fig. 7(b), using the measured gain factor and HITRAN parameters.Due to the high sensitivity of the diagnostic, a small amount of non-zero background absorption was detected near time zero due to broadband absorption by n-octane.In nonreactive n-octane/N 2 experiments, the background absorption was found to be approximately constant and, therefore, the measured A CEAS signal was adjusted with a zero offset.Despite the highly challenging mechanical environment, the CEAS laser diagnostic provided reliable measurements over a time interval of 100 ms, indicating a robust optical setup.
The rise in CO mole fraction near 25 ms in Fig. 7(b) is due to first-stage ignition (or partial oxidation) of n-octane.Overlaid on the measured CO mole fraction profile in Fig. 7(b) is the simulated CO profile using the LLNL kinetic model for n-alkanes [17] in Chemkin software.The simulation was performed using the same initial conditions as the experiment, with a non-reactive RCM volume profile imposed to model the heat loss after compression.

Fig. 1 .
Fig. 1.A schematic of the cross-section of the RCM combustion chamber showing the setup of the optical cavity and off-axis laser alignment.

Fig. 2 .
Fig. 2. (a) Laser intensity traces for an on-axis cavity (red) and Ge etalon (black) for 500 ns laser pulse at 100 kHz repetition rate; (b) Spectral tuning range of the QCL pulse for different pulse duty cycles and repetition rates.

Fig. 3 .
Fig. 3. Measured off-axis laser intensity traces for different pulse durations at 100 kHz repetition rate.

Fig. 4 .
Fig. 4. Variation of (a) chirp rate and (b) laser line-width across the laser pulse for different pulse durations at 100 kHz repetition rate.

FSRFig. 5 .
Fig. 5. Two coupling mechanisms between the laser emission mode (LEM) and the off-axis cavity mode (OCM): (a) rapid frequency down-chirp and (b) broad laser line-width relative to cavity FSR.