Open Access
Add to BibSonomy
Abstract
Molecular oxygen (O2) concentration is measured by employing nanosecond laser-induced plasmas (ns-LIP) over a broad temperature spectrum ranging from 300 K to 1000 K, in the presence of an additional oxygen-containing molecule, CO2. Typically, emission spectra emanating from ns-LIP are devoid of molecular information, as the ns-LIP causes the dissociation of molecular species within the plasma. However, atomic oxygen absorption lines that momentarily appear at 777 nm in the broadband emission from the early-stage plasma are determined to be highly sensitive to the O2 mole fraction but negligibly affected by the CO2 mole fraction. The atomic O absorbing the plasma emission originates from the O2 adjacent to the plasma: robust UV radiation from the early-stage plasma selectively dissociates adjacent O2, exhibiting a relatively low photodissociation threshold, thus generating the specific meta-stable oxygen capable of absorbing photons at 777 nm. A theoretical model is introduced, explicating the formation of the meta-stable O atom from adjacent O2. To sustain the UV radiation from the plasma under high-temperature and low-density ambient conditions, a preceding breakdown is triggered by a split laser pulse (532 nm). This breakdown acts as a precursor, seeding electrons to intensify the inverse-Bremsstrahlung photon absorption of the subsequent laser pulse (1064 nm). Techniques such as proper orthogonal decomposition (POD) and support vector regression (SVR) are employed to precisely evaluate the O2 mole fraction (<1% uncertainty), by analyzing the short-lived (<10 ns) O2-indicator depicted in the early-stage plasma.
© 2023 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Laser-induced breakdown spectroscopy (LIBS) functions as an optical diagnostic technique, facilitating remote, non-intrusive, and precise gas property assessments in demanding test environments. Conventionally, LIBS utilizes focused nanosecond (ns) laser pulses to generate plasmas. Multi-photon ionization in the focal volume initially produces free electrons that swiftly absorb the laser photons during the laser radiation period via the inverse-Bremsstrahlung (IB) process. This accelerates the cascade ionization (CI), causing the dissociation of molecular species in the plasma [1]. The light emission emanating from this plasma, conveying data on its atomic elements, is spectrally resolved, translating into key attributes such as atom composition and gas mixture density. Local equivalence ratio and density in combustion processes [2–7], wherein conventional sensors are rendered ineffective, can be determined by the ns-LIBS. Nonetheless, ns-LIBS cannot provide molecular information [8,9] owing to the IB-induced rapid dissociation of molecules, an issue long considered an inescapable limitation of ns-LIBS. For instance, the O emission intensity from ns-LIP was confirmed to be merely a function of atomic O concentration, independent of its parent molecules, including CO2, O2, and H2O or their mixtures [10].
The need for local O2 monitoring within and at the exit of the combustors has become increasingly critical as recently developed environmentally friendly and high-power combustion systems operate under fuel-lean and highly turbulent flow conditions; the concentration of O2 can serve as an indicator of combustion efficiency. A promising method for optically gauging O2 concentration is diode laser absorption spectroscopy [11–14], providing remarkable accuracy for two-dimensional (2D) flow fields. However, its line-of-sight nature renders it unsuitable for highly three-dimensional systems, such as scramjets/ramjets and swirl-stabilized turbulent combustors, where point measurement methods are favored. Coherent anti-Stokes Raman spectroscopy (CARS), though one of the most precise 0-D composition measurement techniques, requires three beams from at least two different laser sources intersecting at a point, rendering it difficult to maintain optical alignments with vibrating engines at elevated ambient temperatures. Ns-LIBS was demonstrated to quantify gas properties with a minimal optical setup: one laser beam and a single optical access port in realistic test conditions [15].
In this research, a short-lived O2-indicator is newly identified in the early-stage ns-LIP spectra and examined for localized O2 concentration monitoring. The O2-indicator consists of O-absorption lines at 777 nm superimposed on the broadband emission, only appearing within 10 ns after the onset of the breakdown. These lines are revealed to be highly sensitive to O2 mole fraction (XO2). The absorption line strength and width are calibrated employing a conventional fitting method [16–18] for comparison, and proper orthogonal decomposition (POD) is conducted for enhanced calibration accuracy [19,20]. When combined with conventional ns-LIBS (recording the plasma emission spectra after the laser radiation period), simultaneous measurements of O2 concentration, atom composition, and gas density become feasible.
2. Theoretical model
A theoretical model delineating the interaction between the ns-LIP and the adjoining gas mixture is introduced to enhance the comprehension of the observation and furnish a foundation for the analysis of experimental outcomes. The ultraviolet (UV) radiation emitted from the ns-LIP, occupying 1 mm3 volume for the laser pulse duration (<10 ns), possesses sufficient intensity to dissociate O2 in close proximity to the plasma, while its strength rapidly diminishes away from the focal point. Within this volume, i.e., in the absorption layer surrounding the plasma (Fig. 1), oxygen molecules dissociate into metastable oxygen atoms, denoted as O (3s 5S). Comprehensive dissociation pathways resulting in 3s 5S O via UV radiation, such as direct dissociation from super-excited O2 (3Πu) and cascade from high quintet atoms, were previously reported [21–23].
Fig. 1. Schematic diagram of the absorption layer formed around the ns laser-induced plasma, and O (3s 5S) forming mechanism.
The lifetime of the metastable 3s 5S O is relatively extended, causing its population to increase during the cascade process, subsequently leading to the rapid absorption of photons at 777 nm to excite the 3s 5S O. Specifically, the transition absorbing the visible 777 nm photons is readily detectable and does not coincide with other robust transitions, which is advantageous for the quantifiable property (O2 concentration) indicator. The metastable O (3s 5S) generated through UV radiation primarily originates from O2 and not from other O-containing molecules such as CO2. The O-concentration production rate via the dominant 3s 5S O formation pathway is written as follows:
Given that the absorption line intensity at 777 nm is proportional to [O], Eq. (2) conveys that the 777 nm absorbance serves as a direct indication of the O2 concentration in proximity to a robust UV source, such as the early-stage plasma examined in this study.
3. Experimental setup
The experimental apparatus, as illustrated in Fig. 2, comprises three principal components for (1) supplying temperature-regulated gas mixtures, (2) initiating optical breakdown, and (3) recording the plasma emission spectrum. Two pre-prepared gas combinations, N2 (X = 79%)-O2 (X = 21%) and N2 (79%)-CO2 (21%), are blended to modify XO2 from 0% to 21% while maintaining the concentration of N and O constant; that is, the number of atoms per unit volume remains unaltered. The flow rates of these two mixture streams are independently regulated by mass flow controllers, and the aggregate flow rate is fixed at 9 standard liters/min (SLPM). A static mixer is employed to achieve homogeneous mixing of the two streams. Subsequently, an inline gas heater (HT) with an outlet diameter of 13 mm alters the mixture temperature (T) to range from 300 K to 1000 K.
Fig. 2. Schematic of the experimental setup comprising three components for supplying gas mixtures, inducing optical breakdown, and capturing plasma emission spectrum.
A dual-pulse (DP) optical breakdown configuration, as recommended in Refs. [24–27], is employed to sustain robust plasma emission across the extensive temperature range T. Visible (VIS, 532 nm, 30 mJ) and infrared (IR, 1064 nm, 150 mJ) beams, divided from a single ns-laser pulse emanated from a Nd:YAG laser system (Powerlite 8000, Continuum) with an injection seeder intersect at 45° on a horizontal plane 2 mm above the heater outlet. The VIS beam is focused by a plano-convex lens (FL2, f = 75 mm) at the intersection to engender a plasma for seeding electrons that subsequently absorb photons from the IR arriving approximately 8.7 ns later [28]; the short inter-pulse delay between 5 to 10 ns maximizes the IR photon absorption. The IR beam is focused on the intersection by a plano-convex lens (FL1, f = 500 mm), where the electrons, seeded by the preceding breakdown, accelerate the IR photon absorption through the IB process. The time-averaged laser pulse energy following the photon absorption by the plasma is measured using thermal-type power meters (PM) to monitor the plasma energy, thereby governing the intensity of the plasma emission.
The plasma emission is collected from a direction orthogonal to the IR beam path. The absorption line profile of the broadband plasma emission is captured before the formation of the shockwave induced by the breakdown. The plasma emission is channeled to a 30 μm-wide entrance slit of the spectrometer (AM-510, Acton Research Corp.) via a camera lens (CL, Tamron 35–150 mm F/2.8–4 Di VC OSD) and a pair of broadband mirrors (PSM). Subsequently, a sCMOS camera (dicam c1, PCO) positioned at the spectrometer’s exit plane captures the dispersed plasma emission over a 4 ns period starting from the IR beam’s arrival at the focus. This spectrometer-camera combination provides a spectral resolution of 0.0087 nm/pixel, fine enough to resolve the 777 nm O I triplet structure. Fifty shots are documented for each test case. To relocate the plasma relative to the heater’s exit plane and assess the spatial resolution of the measurement technique, the heater is installed on a translational stage (accuracy < 10μm).
In analyzing the spectrum, proper orthogonal decomposition (POD) is adopted. The POD performs singular value decomposition of the covariance matrix consisting of the collected LIBS spectra to extract POD bases. The bases are the most O2-sensitive spectral features since the spectra were collected varying O2 mole fraction in a wide range. Every spectrum is projected onto the space spanned by the bases, and the projection coefficients of each spectrum, i.e., scores, quantify the contributions of the bases to the spectrum. Then the combination of the scores represents the entire spectrum profile, particularly the features sensitive to the O2 concentration. Therefore, the score space remarkably reduces the dimension of the spectrum data for convenient training of the support vector regression (SVR) model. The SVR model is used to estimate the O2 mole fraction taking an absorption spectrum, represented by the POD scores, as the input. More detailed information on the POD spectrum analysis procedure can be found in our previous articles [19,20].
4. Results and discussion
4.1 Effect of dual pulse
Two laser pulses, separated from a single ns-laser pulse, are concentrated at a location with a temporal gap of 8.7 ns. The initial laser pulse’s electron seeding considerably amplifies the plasma energy absorption of the succeeding pulse, particularly under low-density conditions corresponding to T > 500 K (Fig. 3). Generally, low-density gas mixtures necessitate augmented laser pulse energy for optical breakdown. Moreover, the elevated breakdown threshold not only postpones the inception of laser-induced plasmas (LIP) [29,30] but also truncates the duration of rapid IB photon absorption, consequently diminishing the plasma energy absorption. For instance, the ns-LIP elicited by a single pulse (SP, 532 nm) of 180 mJ/pulse (black circle, Fig. 3) absorbs 93% of the pulse energy at 330 K but a mere 13% at 1008 K. Conversely, seed electrons emanating from the prior breakdown induced by the VIS pulse (30 mJ) notably curtail the breakdown threshold of the IR pulse and its reliance on the gas density. Consequently, the DP plasma (black diamond, Fig. 3) absorbs 51% of the total pulse energy, consisting of 30 mJ/pulse (532 nm Seed) + 150 mJ/pulse (1064 nm Pump), at 1043 K, which is considerably higher than that of the SP configuration. This enhancement in plasma energy absorption elevates the plasma temperature and electron population, engendering more stronger UV radiation via the blackbody radiation [31], Bremsstrahlung photon emission, and the radiative recombination process [32]. Ultimately, this contributes to a superior signal-to-noise ratio for the O absorption line at 777 nm.
Fig. 3. Plasma energy absorption for single and dual-pulse configurations in a broad T range. Black dashed line indicates the laser pulse energy in the single pulse configuration (532 nm, 180 mJ), or the sum of pump (1064 nm, 150 mJ) and seed energy (532 nm, 30 mJ) in the dual-pulse configuration.
4.2 Effects of molecular oxygen mole fraction and temperature
The raw spectra are shown in Fig. 4. The dips proximate to 777 nm were observed within the broadband emission, with their depths chiefly dependent on the XO2. The transmittance and absorbance of each spectrum were estimated by fitting the baseline using the four points of intersection between the spectrum and four vertical lines (Fig. 4).
Fig. 4. Raw spectra with various XO2 at 311 K near 777 nm. The four vertical lines indicate four locations used for fitting the linear baselines to calculate the transmittance and absorbance.
Figure 5(a) presents the absorption line profiles of the metastable O monotonically strengthening with increasing XO2. The total quantity of O atoms contained within the gas mixture remains constant across all instances; O2 molecules are merely replaced by CO2 molecules. Given that the minimum photon energies required to yield O (3s 5S) from CO2 and O2 are 20.6 eV and 14.5 eV, respectively [22], the metastable O would be predominantly from O2, thereby serving as an indicator of its concentration, as expressed in Eq. (2).
Fig. 5. Changes in (a) spectrum at 311 K and (b) total absorbance at four T values when XO2 changes from 0% to 21%. The error bars indicate the standard deviation.
The total absorbance of the triplet O absorption line, defined as the integrated absorbance over the wavelength range spanning from 776.5 nm to 778.5 nm, is weakly sensitive to the ambient temperature (T), as shown in Fig. 5(b); However, the absorption line shape, as shown in Fig. 6(a), is T dependent; particularly, the linewidth decreases as T increases. This observation suggests that measurements of T may be feasible by exploiting the dependency of the absorption line profile.
Fig. 6. (a) Min-max normalized absorption spectra in air, and (b) the Lorentzian gamma of the triplet O absorption lines with varying O2 concentration (XO2) and ambient temperature (T). The error bars indicate the standard deviation.
To systematically quantify the width of the triplet O absorption line, the absorption profiles are approximated through the superposition of three Lorentzian line profiles. The Lorentzian gamma, in this context, signifies the absorption linewidth that decreases as T increases, as depicted in Fig. 6(b), presumably attributable to the reduced number density of particles colliding with the O atoms. Moreover, a phenomenon of line-broadening with increasing XCO2, consequently decreasing XO2 at a constant T, has been observed. This occurrence is conjectured to be the result of collisional broadening, instigated by the increased number density of the collision partners interacting with the metastable O.
The absence of interference from another major O-containing species in combustion products, namely H2O, with the 777 nm absorption lines was corroborated in a supplementary experiment employing an identical setup. In this specific experiment, the heater was substituted with a heating mantle that was utilized to boil deionized water in a flask. At the flask’s outlet, steam at 100 °C under atmospheric conditions—presumed to be either pure water vapor (XH2O = 1) or a gas mixture with high water content—was generated, and the spectra were captured at a precise time and duration, adhering to the same experimental procedure. The resulting absorbance at 777 nm was minimal, signifying that the absorbance influenced by the water vapor in the combustion products (10–15%) would be indiscernible. According to Freund [33], the dissociation pathway of H2O that leads to the formation of O (3s 5S) encompasses a quintet state, necessitating an unlikely spin alteration of 2 from the singlet ground state.
4.3 Prediction accuracy and spatial resolution
Spectra—encompassing 50 from 10 temperature conditions of 6 mixture-composition cases—are divided into two sets: 2,400 for the training of a calibration model and 600 for the validation of the model. Two primary POD bases, B1 and B2, are extracted from the training set, as depicted in Fig. 7, accounting for 82.4% and 3.2% of the spectrum variance, respectively. These bases symbolize the spectral features sensitive to XO2 and T, which had been previously correlated with the total absorbance (Fig. 5) and the absorption linewidth (Fig. 6), respectively. Notably, B1 and B2 also represent the total absorbance and the linewidth, respectively, as the training set is collected by altering these specific properties. Any spectrum within the training set can be fully replicated by linearly combining the two bases, resulting in a mere 1.4% relative error.
Fig. 7. Raw spectra with water vapor (XH2O = 1) and room air near 777 nm from a supplementary experiment.
The two coefficients of the bases in the linear combinations, utilized to reconstruct the training set spectra, are illustrated in Fig. 8; here, Score1 and Score2 denote the POD coefficients of the bases B1 and B2, respectively. Specifically, Score1 decreases with an increase in XO2 (Fig. 8(a)), and Score2 decreases with a rise in T (Fig. 8(b)). As observed in the plots, a subdivision of constant XO2 is separable in the score space (Fig. 9(a)), and T (or gas density) could possibly be extracted by partitioning the score space according to the T distribution. Nevertheless, the accuracy of the T estimation employing the score space is likely to be constrained, primarily due to the overlapping of the isothermal subdivisions, particularly in instances where XO2 is low, and Score1 is high (Fig. 9(b)).
Fig. 8. Two major POD bases extracted from the training set, each accounting for 82.4% and 3.2% of the spectrum variance.
Fig. 9. Scatter plots of the training spectra in the score space with the colormaps representing (a) molecular oxygen mole fraction (XO2) and (b) temperature (T), respectively.
A support vector regression (SVR) model is meticulously trained with a set of 2,400 spectra for the refined estimation of O2 concentration, employing the POD-decomposed O absorption line profiles. The 2D score space depicted in Fig. 9 is defined by the two most property-sensitive directions, namely the bases, and all spectra decomposed by these bases are strategically positioned within this space. A significant feature of this method is the substantial reduction in the dimensionality of the spectrum data, a process facilitated by the POD, presenting the primary directions of spectrum variation in response to alterations in properties XO2 and T. The SVR model subsequently establishes a correlation between the combination of the two scores (that is, the location of a spectrum in the 2D score space) and the property of interest, XO2. The spectra encompassed in the validation set, which are not utilized during the model training, are decomposed by the bases corresponding to the training set spectra and distributed within the score space (as denoted by the black dots in Fig. 10(a)). As shown in Fig. 10(b), utilizing the scores corresponding to the validation set spectra, the SVR model well predicts XO2. The relative error in this prediction is restricted to less than 4% (< 1% in XO2 in every instance, accompanied by negligible bias. The accuracy of this XO2 prediction is facilitated by the effective separation of the impact arising from the T variation, employing the second basis B2. For the practical application of the local O2 concentration measurement method, an assessment of the spatial resolution of the technique is conducted. A faster jet consisting of N2 (79%)-CO2 (21%) mixture, devoid of O2, at 18 SLPM is injected into ambient air (N2 (79%)-O2 (21%)). This creates a distinct boundary where XO2 abruptly rises at the jet boundary; the plasma is located merely 2 mm downstream of the jet exit plane.
Fig. 10. (a) Scatter plot of the validation set spectra (black dots) in the score space overlapped on the calibration data, and (b) XO2 predicted by the SVR model considering the spectrum scores as the input.
As illustrated in Fig. 1, the elongation of the absorption layer would increase the measurement volume applicable to this technique. In Fig. 11, the plasma is moved from the jet’s center toward the jet boundary at r = 6.5 mm, oriented perpendicular to the path of the laser beam. The plasma exhibits axisymmetry but is elongated along the beam path. Upon contact of the absorption layer with the ambient air outside the jet boundary, the normalized Score1 commences a decrease from 1 (XO2 = 0%) down to -1 (XO2 = 21%), where the absorption layer surrounding the plasma is situated in ambient air away from the jet boundary. Consequently, the diameter of the ring-shaped absorption layer on a cross-section orthogonal to the laser beam path can be approximated by the thickness of the region wherein Score1 transitions from 1 to -1.
Fig. 11. Radial (r-axis) distribution of the normalized Score1 (1: the CO2-N2 mixture, -1: dry air) at T = 306 K and 1019 K. The error bars indicate the standard deviation.
The Score1 profile well fits to a tangent-hyperbolic function, and in the 5–95% interval of Score1 defined as the diameter of the absorption layer, the profiles are discerned to be 1.05 mm and 1.32 mm at 306 K and 1019 K, respectively. The true spatial resolution would likely be smaller as the diffusion and viscous mixing layer on the jet boundary could enlarge the region of Score1 variation. The cylindrical plasma volume will only shift the absorption layer location, within 0.5 mm in this case, while not affecting the measurement resolution.
This measurement also confirms that the thin (less than a few hundred microns) absorption layer covering the plasma (Fig. 1) is accountable for the 777 nm photon absorption.
5. Conclusion
This research investigates the applicability of conventional ns-LIBS for rendering molecular information, with an emphasis on the detection of molecular oxygen in high-temperature environments. Ns-laser-induced plasmas (ns-LIP) have demonstrated efficacy in diagnosing fuel–air ratios in diverse combustors and combustion processes. However, a challenge arises as these ns-LIPs dissociate molecules within the plasma, thereby promptly the elimination of any trace of the parent molecules. In the current study, a novel molecular oxygen (O2) indicator is identified during the brief period of ns-laser radiation, particularly the O triplet absorption line at 777 nm, manifested on the broadband emission spectrum of the early-stage plasma. The metastable O that absorbs the 777 nm radiation is generated in a thin layer encircling the plasma (originating from nascent O2) via UV photo-dissociation. This new O2-indicator was detected within a 4 ns exposure subsequent to the arrival of the 1064 nm laser pulse at the focal point, over an ambient temperature range of 300 K to 1000 K. The experimentation varied the O2 mole fraction by O2 with CO2 while maintaining the total quantity of O atoms. The sensitivity of this new diagnostic approach to the presence of O atoms originating from O2 (vs. CO2) demonstrates potential for adapting the ns-LIBS diagnostic to differentiate reactants from products in combustion processes. The investigation employed a dual laser-pulse configuration, utilizing a single laser system, wherein a 532 nm pulse provided electron seeding for the subsequent 1064 nm pulse (arriving 8.7 ns later). This configuration sustained high plasma photon emission intensity under the challenging conditions of high temperature and low density. Additionally, POD effectively separated the temperature’s influence on the absorption line profile and reduced the spectrum data’s dimension for facile calibration. A 2D POD score space was constructed employing two principal POD bases. A SVR model was trained utilizing the POD scores extracted from the training spectrum database, facilitating precise calibration that could predict the O2 mole fraction with less than 1% error. Furthermore, the spatial resolution of this technique has been estimated to be finer than 1.32 mm within the temperature range tested in this study.
Funding
National Research Foundation of Korea (2021R1A2C2012697, 2021R1A4A1032023); Air Force Office of Scientific Research (FA2386-20-1-4054); Agency for Defense Development (UD210034SD).
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. H. Do, “Laser-Induced Breakdown Spectroscopy,” in Optical Diagnostics for Reacting and Non-Reacting Flows: Theory and Practice (American Institute of Aeronautics and Astronautics, Inc., 2022), pp. 477–538.
2. S. Zhang, X. Yu, F. Li, G. Kang, L. Chen, and X. Zhang, “Laser induced breakdown spectroscopy for local equivalence ratio measurement of kerosene/air mixture at elevated pressure,” Opt. Laser Eng. 50(6), 877–882 (2012). [CrossRef]
3. H. M. Jun, J. H. Kim, S. H. Lee, and J. J. Yoh, “Towards simplified monitoring of instantaneous fuel concentration in both liquid and gas fueled flames using a combustor injectable LIBS plug,” Energy 160, 225–232 (2018). [CrossRef]
4. B. McGann, T. Lee, T. Ombrello, C. D. Carter, S. D. Hammack, and H. Do, “Inlet Distortion Effects on Fuel Distribution and Ignition in Scramjet Cavity Flameholder,” J. Propul. Power 35(3), 601–613 (2019). [CrossRef]
5. D. Zhang, Q. Gao, B. Li, Z. Zhu, and Z. Li, “Instantaneous one-dimensional equivalence ratio measurements in methane/air mixtures using femtosecond laser-induced plasma spectroscopy,” Opt. Express 27(3), 2159–2169 (2019). [CrossRef]
6. J. Lee, C. Bong, J. Yoo, and M. S. Bak, “Combined use of TDLAS and LIBS for reconstruction of temperature and concentration fields,” Opt. Express 28(14), 21121–21133 (2020). [CrossRef]
7. J. Lee, B. McGann, S. D. Hammack, C. Carter, T. Lee, H. Do, and M. S. Bak, “Machine learning based quantification of fuel-air equivalence ratio and pressure from laser-induced plasma spectroscopy,” Opt. Express 29(12), 17902–17914 (2021). [CrossRef]
8. R. Sattmann, I. Monch, H. Krause, R. Noll, S. Couris, A. Hatziapostolou, A. Mavromanolakis, C. Fotakis, E. Larrauri, and R. Miguel, “Laser-induced breakdown spectroscopy for polymer identification,” Appl. Spectrosc. 52(3), 456–461 (1998). [CrossRef]
9. E. D. McNaghten, A. M. Parkes, B. C. Griffiths, A. I. Whitehouse, and S. Palanco, “Detection of trace concentrations of helium and argon in gas mixtures by laser-induced breakdown spectroscopy,” Spectrochim. Acta, Part B 64(10), 1111–1118 (2009). [CrossRef]
10. B. McGann, C. D. Carter, T. Ombrello, and H. Do, “Direct spectrum matching of laser-induced breakdown for concentration and gas density measurements in turbulent reacting flows,” Combust. Flame 162(12), 4479–4485 (2015). [CrossRef]
11. M. Kroll, J. A. McClintock, and O. Ollinger, “Measurement of Gaseous Oxygen Using Diode-Laser Spectroscopy,” Appl. Phys. Lett. 51(18), 1465–1467 (1987). [CrossRef]
12. L. C. Philippe and R. K. Hanson, “Laser-Diode Wavelength-Modulation Spectroscopy for Simultaneous Measurement of Temperature, Pressure, and Velocity in Shock-Heated Oxygen Flows,” Appl. Opt. 32(30), 6090–6103 (1993). [CrossRef]
13. V. Weldon, J. O’Gorman, J. J. PerezCamacho, and J. Hegarty, “Oxygen sensing using single frequency GaAs-AlGaAs DFB laser diodes and VCSELs,” Electron. Lett. 32(3), 219–221 (1996). [CrossRef]
14. V. Weldon, J. O’Gorman, J. J. PerezCamacho, D. McDonald, J. Hegarty, J. C. Connolly, N. A. Morris, R. U. Martinelli, and J. H. Abeles, “Laser diode based oxygen sensing: A comparison of VCSEL and DFB laser diodes emitting in the 762 nm region,” Infrared Phys. Technol. 38(6), 325–329 (1997). [CrossRef]
15. B. McGann, C. D. Carter, T. M. Ombrello, S. Hammack, T. Lee, and H. Do, “Gas property measurements in a supersonic combustor using nanosecond gated laser-induced breakdown spectroscopy with direct spectrum matching,” Proc. Combust. Inst. 36(2), 2857–2864 (2017). [CrossRef]
16. H. Amamou, A. Bois, B. Ferhat, R. Redon, B. Rossetto, and M. Ripert, “Correction of the self-absorption for reversed spectral lines: application to two resonance lines of neutral aluminium,” J. Quant. Spectrosc. Radiat. Transfer 77(4), 365–372 (2003). [CrossRef]
17. R. Hai, Z. He, X. Yu, L. Sun, D. Wu, and H. Ding, “Comparative study on self-absorption of laser-induced tungsten plasma in air and in argon,” Opt. Express 27(3), 2509–2520 (2019). [CrossRef]
18. K. Touchet, F. Chartier, J. Hermann, and J. B. Sirven, “Laser-induced breakdown self-reversal isotopic spectrometry for isotopic analysis of lithium,” Spectrochim. Acta, Part B 168, 105868 (2020). [CrossRef]
19. T. Yoon, Y. Kang, S. Kim, Y. Park, K. Yee, C. D. Carter, S. D. Hammack, and H. Do, “Proper orthogonal decomposition of continuum-dominated emission spectra for simultaneous multi-property measurements,” Energy 254, 112789 (2022). [CrossRef]
20. S. Kim, J. Choi, H. Park, T. Yoon, and H. Do, “Gas composition measurements in randomly distributed and fast moving gas bubbles in two-phase fluids,” Opt. Express 31(9), 14255–14264 (2023). [CrossRef]
21. R. S. Freund, “Dissociation by Electron Impact of Oxygen into Metastable Quintet and Long-Lived High-Rydberg Atoms,” J. Chem. Phys. 54(7), 3125–3141 (1971). [CrossRef]
22. J. W. McConkey, C. P. Malone, P. V. Johnson, C. Winstead, V. McKoy, and I. Kanik, “Electron impact dissociation of oxygen-containing molecules–A critical review,” Phys. Rep. 466(1-3), 1–103 (2008). [CrossRef]
23. Y. Zhou, Q. Meng, and Y. Mo, “Photodissociation dynamics of superexcited O2: Dissociation channels O(5S) vs. O(3S),” J. Chem. Phys. 141(1), 014301 (2014). [CrossRef]
24. A. P. Yalin, N. Wilvert, C. Dumitrache, S. Joshi, and M. N. Shneider, “Laser plasma formation assisted by ultraviolet pre-ionization,” Phys. Plasmas 21(10), 103511 (2014). [CrossRef]
25. A. Alberti, A. Munafò, C. Pantano, J. B. Freund, and M. Panesi, “Collinear dual-pulse laser optical breakdown and energy deposition,” J. Phys. D Appl. Phys. 53(20), 205202 (2020). [CrossRef]
26. Y. Park, K. Kim, S. Kim, C. D. Carter, and H. Do, “Nanosecond laser pulse modulation using seed electrons from cascade ionization induced by inverse-Bremsstrahlung photon absorption,” Opt. Express 30(4), 6037–6050 (2022). [CrossRef]
27. S. Kim, J. Choi, H. Byun, T. Yoon, and H. Do, “Stabilization of Laser-induced Plasma on Turbulent Flame Boundaries with Electron Seeding,” in Photonics Conference (IEEE, 2022), 1–2.
28. P. K. Diwakar and D. W. Hahn, “Study of early laser-induced plasma dynamics: Transient electron density gradients via Thomson scattering and Stark Broadening, and the implications on laser-induced breakdown spectroscopy measurements,” Spectrochim. Acta, Part B 63(10), 1038–1046 (2008). [CrossRef]
29. J. Kiefer, J. W. Tröger, Z. Li, T. Seeger, M. Alden, and A. Leipertz, “Laser-induced breakdown flame thermometry,” Combust. Flame 159(12), 3576–3582 (2012). [CrossRef]
30. S. Oh, C. D. Carter, Y. Park, S. Bae, and H. Do, “Non-intrusive laser-induced breakdown spectroscopy in flammable mixtures via limiting inverse-bremsstrahlung photon absorption,” Combust. Flame 215, 259–268 (2020). [CrossRef]
31. A. Borghese and S. S. Merola, “Time-resolved spectral and spatial description of laser-induced breakdown in air as a pulsed, bright, and broadband ultraviolet-visible light source,” Appl. Opt. 37(18), 3977–3983 (1998). [CrossRef]
32. A. De Giacomo, R. Gaudiuso, M. Dell’Aglio, and A. Santagata, “The role of continuum radiation in laser induced plasma spectroscopy,” Spectrochim. Acta, Part B 65(5), 385–394 (2010). [CrossRef]
33. R. S. Freund, “Electron-impact dissociation of H2O and D2O into metastable fragments,” Chem. Phys. Lett. 9(2), 135–138 (1971). [CrossRef]
