1. Introduction

Laser ablation (LA) has a vast number of applications in various fields, and plasma diagnostics play a key role in optimization of the ablation process for each application. All diagnostic tools developed for steady state plasmas in the past can essentially be used to diagnose and optimize transient laser plasmas for various applications; however, caution must be taken when interpreting the results by considering the large gradients in plasma parameters—both spatially and temporally [1]. The most widely-used tools for analyzing laser plasmas are optical emission spectroscopy (OES) [2] and electrical probes [3] because of their simplicity, but they come with several assumptions and/or limitations. Some of the limitations inherent to OES are opacity effects of the plasma, free-free and free-bound transitions at early times of plasma evolution that obstruct viewing the line emission originating due to bound-bound transitions, instrumental broadening from the spectrograph and detection system that limits the spectral resolution, errors associated with atomic and molecular fundamental constants, unavailability of spectroscopic constants for high-Z elements, etc. Electrical probe (Faraday cup, Langmuir probe, etc.) measurements require a vacuum environment and fast electronics to gain information with high time resolution [4]. Discrimination between different components of the plasma (different species, different charge states, etc.) is also extremely challenging. Moreover, fast-moving particles can sputter the probe, leading to erroneous results.

Among the various plasma diagnostic tools, active optical diagnostics arguably provide the most accurate results. The most well-known optical diagnostic tools are Thomson scattering, laser absorption spectroscopy (LAS), laser-induced fluorescence (LIF), and interferometry. Thomson scattering and interferometry are, perhaps, the best tools to measure the plasma fundamental properties with highest accuracy. Compared to OES, both LAS and LIF provide selectivity and sensitivity. Moreover, LAS and LIF are capable of resolving the inherent linewidths of a transition and are not limited by instrumental line broadening. However, all optical diagnostic tools require a second laser with certain stringent parameters. For example, high energy pulses are required to obtain scattered signal from electrons, especially to overcome strong Rayleigh scattering. Short pulse lasers (⪝100 ps) are necessary to avoid smearing effects in interference patterns due to large density gradients seen in laser-produced plasmas (LPP) at early times [5]. Narrowband lasers are preferable for performing LAS and LIF, especially to measure inherent linewidths. LAS also has very strict geometry requirements: the laser beam must pass through the plasma plume parallel to the target so that the transmission (and therefore absorbance) can be measured. By measuring the transmission, LAS is also subject to noise from particles and turbulence in the plume, which can make measurements of low-concentration elements at atmospheric pressure challenging [6].

Currently, laser-induced breakdown spectroscopy (LIBS) is a popular application of LA combined with OES due to its simplicity, robustness, standoff capability, real time analysis with no sample preparation requirement, minimal sample destruction, and the possibility of multi-elemental analysis [7–10]. However, the analytical merits (detection sensitivity, precision, etc.) of LIBS are not as strong as other commonly used analytical methods (e.g., mass spectrometry) for trace elemental and isotopic detection. Moreover, the spectral features from targets containing high-Z elements are very crowded [11] and sometimes interpreting the line features is a challenge, even though chemometrics helps to overcome some of these limitations [12]. Recently, extensive research has been carried out to improve the figures of merit of LIBS; some methods are microwave-excited LIBS, double pulse LIBS, cavity-enhanced LIBS, spark excitation, and resonance-enhanced LIBS [8, 13]. These methods typically improve the signal-to-noise (S/N) ratios of the LIBS measurement; however, they all come with various other complexities. LIF is capable of addressing some of the limitations of the figures of merit of LIBS (precision, detection limits, line broadening, isotopic analysis, etc.). However, in spite of its several advantages, LIF has not been routinely used to diagnose transient laser plasmas or to extend LA applications to various fields. It has to be mentioned that along with sensitivity, LIF is tuned to certain transitions in the ablation plume (selectivity). Combining LIF and LIBS allows us to broaden the scope of LA-based analytical methods with multi-elemental detection capability and high sensitivity.

Limited studies are available in the literature for LIF of LA. LIF of LA plumes was first reported by Measures and Kong—they found improved detection limits, reduced matrix effects, and a linear calibration curve in comparison with LIBS [14, 15]. In their experiment, a nitrogen laser-pumped tunable dye laser was tuned to the transition wavelength of an analyte of interest in an LPP plume. Niemax et al. used a pulsed dye laser for LIF and demonstrated calibration of LIF data by internal standardization with metallic and gas matrices [16]. They extended this work using a continuous-wave (cw) diode laser for LIF excitation and reported uranium isotope ratios in solid samples [17]. In the past, LIF of LA plumes was used for various trace detection applications which include the determination of Pb in metallic reference materials [18]; Pb in aqueous solutions [19]; Co in soil, steel and graphite [20]; isotopic analysis of Li [21]; heavy metals in soils [22]; trace boron in nickel-based superalloys and steels [23]; Al, Cr, Fe and Si in steel [24]; Pb in a Cu matrix [25]; etc.

Miyabe et al. investigated the dynamic behavior of LA plumes in the presence of an ambient gas of 6 Torr He using LIF and reported that a significant portion of ground state atoms and ions accumulate in the contact region between the plume and ambient gas [26, 27]. Simultaneous optical absorption and LIF measurement were used for 3D mapping of the number densities of ions and neutrals in a laser ablation plume in vacuum by scanning the pulsed dye laser frequency [28, 29]. Such studies are extremely useful for evaluating the fundamental properties as well as the hydrodynamic expansion features of LA plumes. Several research groups explored resonance-enhanced LIBS (RELIBS) analysis for trace metal detection and Pb in an aqueous solution [19, 25, 30–33]. In RELIBS, the probe laser is used to resonantly excite the host atoms, and energy exchange happens through collisional transfer. Nakata et al. studied quenching effects during LIF of LA plumes using 2D photography [34]. In another study, they evaluated the spatially- and temporally-resolved evolution of ablated atoms for different ablation laser energies and ambient conditions [35] and extended their studies to gain an understanding of molecular formation during pulsed laser deposition [36]. Orsel et al. employed LIF to study oxidation of Al in LA plumes using spatio-temporal mapping [37]. LIF was previously used to image nanoparticles in LA plumes with the help of an atomization laser beam [38].

Fluorescence spectroscopy is routinely used to probe molecular structure and dynamics [39], in biomedical applications [40], for environmental sensing [41], and in combustion research [42]. The general parameters of fluorescence spectroscopy for probing and sensing are the excitation spectrum, the emission spectrum, and the fluorescence lifetime or decay rates. 2D-fluorescence spectroscopy (2D-FS), which covers a wide range of excitation and emission wavelengths, is a relatively new technique and has been previously used to probe actinide chemistry [43], to visualize the equilibration dynamics of complex molecules with high time resolution at the single-molecule level [44], and to monitor biological cell growth and metabolic changes [45]. Recently we demonstrated the use of 2D-FS for the first time in highly transient LPPs by measuring simultaneous absorption and emission from coupled electronic transitions of Al in air [46]. 2D-FS of LPPs allows the measurement of fluorescence emission spectra over a wide range of excitation wavelengths and may extend the applications of LPPs into new areas such as precise isotopic analysis, measurement of inherent linewidths of atomic and molecular emission transitions, increased understanding of nucleation and condensation in LPPs, etc.

Even though LIF is a very powerful technique that is regularly used in other research areas (biomedical, combustion, remote sensing of gas phase, etc.), quantitative applications of LIF are limited mostly because of the lack of a predictive understanding of LIF signal generation as well as its quenching, especially in complex environments like LPPs. An LPP is a high-density collective system; its properties change rapidly with space and time, and hence collisional conditions in the plasma plumes alter both the excited and ground state level populations along with spectral line shapes. Since LIF signal intensity is particularly sensitive to various electronic energy transfer rates (quenching rates), significant LIF signal intensity variation can be anticipated with changes in local conditions (ambient gas medium, pressure, etc.). Some of the quenching mechanisms that are relevant during laser-plasma expansion, especially in the presence of an ambient gas, include excited-state reactions, energy transfer, ground state complex formation (chemical), and collisional quenching [47].

In this article, we use a cw laser as an ultra-narrowband tunable LIF excitation source to increase the magnitude and persistence of emission from selected atomic transitions of U and Al species. We specifically investigate the role of ambient gas pressure on LIF signal intensity and persistence. The temporal features of LIF emission are evaluated and compared with the LIBS signal. 2D-FS absorption/emission images are recorded at various nitrogen pressure levels, showing enhanced emission from the selected transitions. The excitation spectral features showed the inherent linewidths of various transitions in the plasmas, which are significantly narrower than observed in emission spectra. The present results also highlight the advantages of LIF as a diagnostic tool employing cw laser re-excitation to overcome some of the well-known limitations of traditional emission and absorption spectroscopic methods.

2. Experimental details

A schematic of the experimental setup is given in Fig. 1. The fundamental radiation from a Q-switched Nd:YAG laser with a pulse width of ~6 ns (full width half maximum, FWHM) and a repetition rate of 10 Hz was used to produce the plasmas. The laser energy at the target surface was varied by a combination of a waveplate and a cube polarizer. A fixed laser energy of ~17 mJ was used to produce the plasma, and the corresponding power density at the target was 2 GW.cm⁻². The laser was focused onto a glass target (Kopp 3750 filter glass) containing 1.3% ²³⁸UO₂ and approximately 1-2% Al₂O₃ by weight. Ablation was performed in a small chamber with quartz windows. The ablation chamber was mounted on an x-y-z translation stage which was moved to avoid target cratering effects as well as to align the excitation beam with respect to the plasma plume. A cw, frequency-doubled Ti:Sapphire laser (M-Squared SolsTiS) was used to excite the ground state population in the plasma. The laser is tunable in the spectral range of 350-500 nm and provides an ultra-narrow linewidth output (⪝100 kHz) with cw power levels of ~300 mW and a beam diameter of ~1 mm. The wavelength of the excitation laser was monitored using a wavelength meter. The LIF excitation laser was passed parallel to the target surface for re-excitation without any focusing optics and was ~1 mm from the target surface. The LIF measurements were performed at various ambient pressures from ~3 Torr to atmospheric pressure levels of N₂. Ultrahigh purity grade (99.999%) N₂ gas was used, and the N₂ flow rate was varied to maintain a constant pressure within the chamber.

 

Fig. 1 The schematic of the experimental setup (WP, waveplate; C, polarizing cube; BD, beam dump; L, lens; DPSS, diode pumped solid state laser; TiSa, Titanium Sapphire laser; M, mirror; TG, timing generator; ICCD, intensified charge-coupled device).

Download Full Size | PDF

To collect the LIF signal, a plano-convex lens with focal length 5.0 cm and aperture 2.2 cm was positioned at an angle ~25 degrees with respect to the target normal; the collected light was transported using a 400 µm multimode fiber optic cable. The other end of the fiber was connected to a 0.5 m spectrograph (Acton, Spectrapro). The spectrograph consists of three gratings (150 g/mm, 600 g/mm and 2400 g/mm) and two detectors: a photomultiplier tube (PMT, R928, Hamamatsu, 2 ns rise time) for single channel detection and an intensified CCD (ICCD, PiMax, Princeton Instruments) for multichannel detection. The combination of the 2400 g/mm grating in the spectrograph and ICCD provided a resolving power >10,000. All spectral measurements were performed in a spatially-integrated manner.

3. Results and discussion

LIF of LA plumes was performed employing different methods. Initially, a time evolution study of both LIBS and LIF was conducted to evaluate the temporal dependence of thermally-excited (LIBS) and LIF photons. Then, a time-integrated LIF measurement was carried out using a proper time delay and gate width to exclude LIBS emission for various excitation wavelengths and different ambient nitrogen pressures. These data were used to generate 2D-FS absorption/emission maps. The spectral lines selected for LIF excitation were the 394.3816 nm (0-25,348.97 cm⁻¹) U transition and 394.4006 nm (0-25,347.756 cm⁻¹) Al transition, both resonance lines. The Grotrian diagrams of the selected transitions are given in Fig. 2. The frequency-doubled Ti:Sapphire laser was tuned across the transition of interest for excitation. The selection of these lines was based on the fact that they are separated by a small spectral region (~18 pm) as well as both the Al and U transition share a common upper level with other transitions (396.15 nm for Al and 404.28 nm and 580.21 nm for U), allowing for the fluorescence signal from non-resonance transitions to be collected and monitored.

 

Fig. 2 Partial energy diagrams of the U I (left) and Al I (right) transitions selected for the present study are given.

Download Full Size | PDF

3.1 Time-resolved LIBS and LIF emission

Atomic emission from LPPs is highly dynamic in nature, with LIBS emission resulting from thermal excitation of species predominant at early times when the plasma temperature is higher. To distinguish emission resulting from thermal excitation versus resonant laser excitation, time-resolved emission spectra in the spectral region around 404 nm were recorded, which covers a non-resonantly excited U I transition at 404.275 nm and two adjacent K I lines (404.414 nm and 404.721 nm) while the LIF was pumped with the U I transition at 394.3816 nm. Figure 3 shows the 2D emission contours in the absence and presence of the LIF excitation beam at 45 Torr N₂ ambient pressure. The LIF excitation beam was fixed at 394.3816 nm, which shares the same upper energy state with the 404.275 nm transition. To record the temporal evolution of LIBS and LIF emission, the required spectral region was selected and spectra were recorded with 2 µs resolution. As Fig. 3 shows, thermally-excited emission exists only at early times and persists for ⪝15 µs. Compared to U I emission at 404.275 nm, the K I lines at 404.414 nm and 404.721 nm are found be presumably intense because of the higher concentration of the alkali element in the selected glass sample. Among the K I lines, the 404.414 nm transition is more intense compared to 404.721 nm because of its higher gA value. The 2D contours show that in the presence of the resonant excitation beam, the emission from U I is selectively enhanced and persists for a very long time (~100 µs) with maximum LIF emission at ~50 µs. We also recorded the LIF emission features at 580.211 nm. This transition also shares the same upper level with the resonant pumping transition; however, its emission intensity was relatively weak because of an inferior transition probability (A = 2.031 × 10⁶ s⁻¹) compared to the 404.275 nm transition (A = 3.537 × 10⁷ s⁻¹) [48].

 

Fig. 3 Time-resolved emission features from U-containing plumes in 45 Torr N₂ recorded in the absence and presence of the LIF excitation beam. The 2D emission contours were obtained by recording the emission features with a temporal resolution of 2 µs.

Download Full Size | PDF

Since both LIBS and LIF signal intensities are particularly sensitive to various electronic energy transfer rates, significant variation in signal intensity can be expected with changes in ambient pressure. Previous results showed that the nature and pressure of the ambient gas affect the fundamental properties of LA plumes (density, temperature and its evolution, ablation efficiency) [49] along with plume hydrodynamics [50]. With increasing pressure, the plasma becomes more collisional and thermal emission will be enhanced. Previous reports also showed that the optimal pressure conditions for LIBS S/N ratios were around ~100 Torr [11]. At higher pressures, the collisional process will be significantly enhanced because of plume confinement. The 2D contours representing the temporal evolution of LIF emission for various ambient pressure levels are given in Fig. 4. With increasing pressure, because of a rise in confinement, the LIF persistence is considerably reduced along with an increase in persistence of thermally-excited emission. At lower pressure, the ground level reservoir persists for a longer time, while with increasing pressure, because of enhanced confinement, various quenching mechanisms viz. excited-state reactions, energy transfer, and collisional quenching will eventually reduce the LIF signal intensity as well as its lifetime.

 

Fig. 4 2D contours of emission in the presence of the 394.38 nm resonant excitation beam at various pressure levels. The ICCD gains used were 100 for 100 Torr N₂ and 150 for 475 and 760 Torr N₂. The gate width used to record the temporal evolution was 2 µs.

Download Full Size | PDF

3.2 2D-fluorescence spectroscopy

The 2D contours given in Figs. 3 and 4 were recorded with the excitation laser wavelength fixed to the resonance transition. The general parameters of fluorescence spectroscopy for probing and sensing are excitation and emission wavelength along with temporal gating. We employed 2D-FS to monitor the LIF emission features at various pressure levels. 2D-FS, which covers a wide range of excitation and emission wavelengths, is a relatively new technique and has been used previously for biological applications and to probe actinide chemistry [28–30]. We recently demonstrated 2D-FS in laser-produced plasmas by measuring simultaneous absorption and emission from coupled electronic transitions of Al in air [31]. We recorded 2D-FS on plasma plumes at various N₂ pressure levels by scanning the excitation wavelength; results obtained under 45 Torr N₂ ambient pressure are given in Fig. 5. To obtain 2D-FS, the LIF emission spectra were recorded with a 15 µs gate delay and 100 µs gate width during the excitation laser scan. Non-resonant LIF emission in the 404 nm region was recorded while the excitation laser beam was tuned across the 394.3816 nm transition. At 45 Torr N₂ ambient pressure, LIBS emission is found to be negligible for the U I transition in the selected temporal window (Fig. 3(a)). A one-dimensional cross section of the 2D-FS map at a fixed emission wavelength gives the corresponding excitation (or absorption) spectrum, and a one-dimensional cross section at a fixed fluorescence excitation wavelength provides the corresponding emission spectrum. The excitation and emission spectra obtained from the 2D-FS plots are given in Figs. 5(c) and 5(d).

 

Fig. 5 (a) 2D-FS map is given showing LIF of U I and weak thermal emission from K I lines at 45 Torr N₂ pressure. (b) Zoomed-in image of 2D-FS of U I transition. To obtain the 2D-FS map, the emission signal was collected at 404.275 nm while the excitation beam was tuned across the 394.3816 nm transition. (c) The excitation spectrum and (d) the emission spectrum. The circles represent the data points and the smooth curves are Voigt fits.

Download Full Size | PDF

Figures 5(a) and 5(b) show the selective excitation of LIF when the excitation laser is tuned across the absorption resonance. Since a non-resonant transition is selected to monitor LIF, resonant scattering of the LIF pump wavelength is not observed as the laser is tuned across the 394.38 nm U I transition while the emission is collected for 404.275 nm transition. The 2D-FS map given in Fig. 5 gives the absorption and emission signature overlap. Figures 5(c) and 5(d) show the spectral resolution available using LIF and LIBS. The recorded linewidth of the LIBS emission is ~45 pm, which represents the instrumental broadening of the spectrograph used, while the measured linewidth from the LIF excitation spectrum is found to be ~0.94 pm. Based on the reported transition probability of the U I transition at 394.38 nm, the natural linewidth of the transition is 0.0216 pm [50]. Since the LIF measurements are performed 15 µs after the plasma evolution, the Stark broadening effect can be neglected as it is predominant at early times in the LPP evolution, especially ⪝5 µs. However, the spectral linewidth contributions from Doppler broadening and van der Waals broadening due to collisions with neutral species (including molecules) cannot be avoided. At moderate pressure levels (45 Torr), Doppler broadening will be significant compared to van der Waals broadening. The Doppler width of a transition is related to temperature according to the equation ${\text{Δλ}}_{\text{D}}=7.16\times {10}^{-7}\text{λ}\sqrt{\frac{{\text{T}}_{\text{D}}}{\text{m}}}$, where λ is the wavelength of the transition, T_D is the kinetic (Doppler) temperature of the plasma, and m is the atomic mass of the species. The estimated kinetic temperature using the measured linewidth is 2209 ± 70 K. These results indicate that the intrinsic (physical) spectral linewidths of transitions in an LPP can be measured using LIF of LA plumes.

We also recorded 2D-FS maps of the U I 404.275 nm transition at various N₂ pressure levels up to atmospheric pressure conditions; results are given in Fig. 6. The interaction of an LA plume with an ambient gas has been studied extensively in the past [50,51]. Compared to plume expansion in vacuum or at moderate pressure levels, the interaction of an LPP with a background gas is more complex at higher ambient pressure levels due to formation of shock waves, clustering, plume splitting and sharpening, instability formation in the plume-ambient interface, deceleration and confinement of the ablated species, thermalization of the ablated species, diffusion, changes in laser-target and laser-plasma coupling, etc. The temperature and density of the plasma plume also have a strong dependence on the pressure and nature of the surrounding gas. As the figure shows, with increasing pressure the change in intensity for both LIBS and LIF emission peaks as well as broadening of the LIF peaks is apparent. The excitation and LIBS emission spectra obtained from the 2D-FS maps are given in Fig. 7. As the pressure increases, the linewidth of the U I 394.38 nm transition is also increased. The measured linewidths (Voigt, Gaussian and Lorentzian FWHM) at various pressure levels are given in Fig. 8(a). The Doppler (Gaussian) component showed insignificant changes with pressure while the Lorentzian FWHM is found to show an increase in linewidth with pressure. This implies an increase in van der Waals broadening with pressure. However, the linewidth of the emission spectra is ~45 pm regardless of pressure, indicating the emission spectral linewidth is limited by instrumental broadening. Typically, because of confinement, the linewidths of all radiation will be increased due to the Stark effect, but its role in the present measurement is limited as the detection was performed with a gate delay of 15 µs. Figure 8(b) shows the U I emission intensity from LIF-enhanced LIBS and LIBS for various pressure levels. With increasing pressure, the LIF emission is found to be reduced with a gradual enhancement in LIBS emission.

 

Fig. 6 2D-FS maps of U I transitions at various N₂ pressure levels: (a) 100 Torr, (b) 275 Torr, (c) 475 Torr, and (d) 760 Torr. All measurements were performed with 15 µs gate delay and 100 µs gate width.

Download Full Size | PDF

 

Fig. 7 (a) Excitation and (b) emission spectra obtained at various pressure levels. The excitation spectra are fitted with a Voigt function to obtain the linewidths.

Download Full Size | PDF

 

Fig. 8 (a) The measured FWHM from the excitation spectral line shape for various pressure levels is given. A Voigt function was used for fitting; the Gaussian and Lorentzian FWHM components are also given. (b) The LIF and LIBS signal variation with pressure for the U I transition.

Download Full Size | PDF

Figure 7(b) shows that the emission intensity from thermal excitation (K I lines) is significantly increased with pressure in conjunction with a decrease in the magnitude of the LIF signal. However, the selective enhancement in the LIF signal is still significant even under atmospheric pressure conditions. It has to be pointed out that based on the 2D-FS maps given in Fig. 6, the thermal excitation contribution of the U I species is negligible even at atmospheric pressure conditions (Fig. 8(b)) in comparison with the LIF signal. The optimal condition for LIF corresponds to a low density, cooler atomic cloud with a sufficient number of atoms available in the ground level for selective absorption and subsequent fluorescence emission. As Fig. 4 shows, the persistence of LIBS emission is enhanced with pressure due to increased collisions from higher plume confinement. However, it has an opposite effect (decreased persistence) on the LIF signal. The reduction in LIF with increased pressure could be due to several effects, including (1) collisional quenching due to increased excitation/electronic temperature; (2) reduction in excitation beam interaction length with the plasma due to higher confinement; (3) a change in the temporal behavior of the population density in the ground state; etc. The plume morphology/density changes significantly with pressure, which in turn affects the LIF beam fractional interaction volume with plasma. For example, the reported size of the plume under approximately similar conditions was ~6 mm and ~3 mm in diameter at 10 Torr and 760 Torr, respectively [50].

The 2D-FS maps recorded at atmospheric pressure levels show a significant increase in signal noise (Fig. 6). This indicates that random flicker noise increases with pressure [52]. The 2D-FS maps represent the emission spectra recorded at each excitation wavelength in the scan without any averaging or accumulation, and we haven’t performed any digital smoothing of the signal. Moreover, it is anticipated that the fluctuation in signal intensity will be more pronounced when the intensity of the radiation source is higher. However, such a fluctuation was not seen when the LIF signal was maximum, which occurred at lower pressure levels (≤100 Torr). So it can be argued that at lower pressure levels, the excitation beam interacts with a certain-sized atom reservoir where pulse-to-pulse variation in the number density of the analyte is minimal. Previous studies showed that both LIF and LAS provide the highest S/N ratios at lower pressure levels for LPP systems [16]. A differential absorption dual-beam technique was found to reduce noise in LAS, especially at higher pressures [6].

3.3 HR spectroscopy using LIF

The excitation spectral information provided in Figs. 5 and 7 show the powerful nature of LIF of LA plumes to obtain high resolution (HR) spectra. It is well known that HR spectroscopy is required to resolve closely spaced atomic transitions and isotope shifts. The measured LIF linewidth of the U I 394.38 nm transition even at atmospheric conditions (~1.44 pm) is significantly lower than the resolution typically available with laboratory-based spectrographs. We selected two nearby transitions to illustrate the spectral resolution that is achievable using LIF. The transitions selected for this study were U I 394.3816 nm and Al I 394.40 nm, separated by 18.4 pm. To resolve such a splitting using emission spectroscopy, we need spectrographs with resolving power ⪎22,000. For comparison, in the present study the emission spectroscopy was performed using a 0.5 m spectrograph with 2400 g/mm grating, which provides a resolving power of ~10,000. The recorded 2D-FS of the selected U I and Al I transitions is given in Fig. 9. The measurement was performed in 100 Torr N₂ pressure by scanning the excitation beam through both resonance transitions, and the resonant emission was recorded. Since there is a higher concentration of Al in the sample and hence a higher number density of Al in the plume, a gate delay of 30 µs was selected along with a gate width of 40 µs to avoid signal saturation. The excitation spectrum obtained from 2D-FS is given in Fig. 9(b), which shows that the lines are well separated. The measured linewidths of the U I and Al I transitions are 0.9 pm and 3.1 pm, respectively. The thermally-excited emission spectrum given in Fig. 9(c) shows the emission from two Al transitions that share a common upper level, with the excitation beam at resonance with the Al I 394.4 nm transition. Figure 9(c) also gives the emission spectrum when the excitation beam was in resonance with the U I 394.3816 nm transition. Figure 9(d) shows the emission spectra when the excitation beam was in and out of resonance with the U I and Al I transitions and highlights the limitation of LIBS for high resolution studies. The present studies showed that by using LIF on LA plumes, isotope shifts of ≲1 pm can be resolved. To obtain such a spectral resolution using emission spectroscopy, spectrographs with extremely high resolving power (λ/Δλ ~400,000) are needed.

 

Fig. 9 (a) 2D-FS of U I and Al I transitions. The measurement was performed at 100 Torr N₂ pressure by scanning the excitation beam through both resonance transitions, and the resonant emission was recorded. A gate delay and width of 30 µs and 40 µs, respectively, were used. (b) Excitation spectrum. (c) Emission spectrum – top panel gives the emission spectrum when the excitation beam was at 394.4 nm; bottom panel – emission spectrum when the excitation beam was at 394.3816 nm U I. (d) Emission spectrum when the excitation beam was at resonance with U I and AI transitions along with non-resonance position (LIBS).

Download Full Size | PDF

The measured LIF linewidth of Al I 394.4 nm (~3.1 pm) is found to be significantly larger compared to the U I transition at 394.3816 nm. The selected Al transition at 394.4 nm (³P_1/2 – ⁴S_1/2) has hyperfine splitting of 1.5 GHz (~0.8 pm) in the ground state (³P_1/2). However, the hyperfine splitting was not resolved in the LIF excitation spectrum. Assuming a Doppler temperature of ~2200 K, the calculated Doppler width of the Al transition is 2.6 pm, which is significantly greater than the hyperfine splitting of the transition. Modeling the convolution of the line based on the calculated Doppler width and the hyperfine structure of the transition [53] results in a total Doppler FWHM of ~2.95 pm, which agrees well with the measured linewidth.

The present studies show the advantages of using a cw laser for LIF excitation over the use of pulsed lasers. Some of the key advantages are the ultra-narrow bandwidth of cw lasers and hence better selective excitation and isotopic analysis, continuous re-excitation of the ground state population during the entire lifetime of the plasma plume, and simpler parametric optimization because of the absence of inter-pulse delay. The present studies also display the advantages of using LIF on LA plumes to overcome some of the inherent limitations of both LIBS and LAS. Our measurements showed that LIF was able to measure the inherent linewidths of various transitions in a transient plume. Since LIF re-excites the ground state population, which persists out to later times when the plasma is cooler, line broadening effects such as Stark broadening, which is predominant in LIBS, will be negligible. The use of a continuous laser also helps to re-excite the ground state population for a prolonged time, leading to enhanced LIF emission at later times of plume evolution. Moreover, since LIF probes laser absorption via detection of emission, it is less subject to noise due to particle emission and turbulence that is observed in transmission-based measurements. Moreover, the demonstrated 2D-FS of LA plumes may be used for various applications of LPPs such as precise measurement of inherent linewidths, isotopic analysis of atomic and molecular emission transitions, and developing a deeper understanding of condensation physics/nucleation.

4. Conclusions

In this article, we report LIF excitation of LA plumes under various ambient pressures up to atmospheric pressure levels. We used a tunable cw laser as an ultra-narrowband source for LIF excitation to increase the magnitude and persistence of emission from selected atomic transitions in LPPs. Time-delayed and gated detection of the emission spectrum was used to isolate the resonantly-excited fluorescence emission from the thermally-excited emission in the plasma. At lower pressure levels (i.e., ⪝100 Torr), LIBS emission dominates at early times (⪝15 µs) and LIF emission persists out to longer times (~100 µs). With increasing pressure, the signal intensity and persistence of LIBS emission is enhanced along with reduction of the LIF signal due to various quenching effects. The reduction in LIF with increased pressure could be due to several effects, which include collisional quenching due to increased excitation/electronic temperature, reduction in excitation beam interaction column length with the plasma due to higher confinement, changes in the temporal behavior of the population density in the ground state, etc.

2D fluorescence spectroscopy (2D-FS) absorption/emission images were recorded at various nitrogen pressure levels, showing enhanced emission from the selected transitions. The 2D-FS maps provided simultaneous absorption and emission overlaps of the selected transitions and are a powerful tool to demonstrate both excitation and emission. The resolution of LIF emission is found to be dictated by the linewidth of the excitation beam rather than the instrumental broadening of the detection system as typically seen in LIBS spectra. The measured linewidth of the LIBS emission features is ~45 pm, which represents the instrumental broadening of the spectrograph and detection system used. The observed linewidth of LIF emission from the excitation spectrum is significantly narrower (⪝1 pm) than LIBS emission and agrees well with estimation of the Doppler width of the transition. The inherent linewidth of the U I transition also increased with increasing pressure, indicating the role of van der Waals broadening in LPPs.

We also evaluated the resolving power of LIF by recording 2D-FS of two spectrally close transitions, separated by 18.4 pm, to demonstrate the effectiveness of excitation spectra over thermally-excited spectra for high resolution studies by resonantly exciting U I and Al I transitions. The measured excitation spectra showed the inherent linewidths of the U and Al transitions. HR spectroscopy is essentially needed to resolve the isotopic shift of atomic transitions, which is typically <10 pm. The present studies showed that by using LIF on LA plumes, isotope shifts ⪝1 pm can be resolved. To obtain such a spectral resolution using emission spectroscopy, spectrographs with extremely high resolving power (~400,000) are needed.

In this paper, we also highlight the advantages of LIF as a diagnostic tool employing cw laser re-excitation over traditional emission and absorption spectroscopic methods. It has to be pointed out that the excitation beam was not focused onto the plasma plume and hence excitation geometry does not play a role here, unlike in LAS where the probe beam is restricted to passing parallel to the target so that transmission through the plume can be detected. It signifies that standoff detection is possible with LIF of LA plumes. However, spurious scattered signal can influence the LIF measurement with an orthogonal configuration. By selecting non-resonant emission, scattered photons can be avoided.