Abstract
A single-beam-splitting approach was employed to enhance the signal intensity of LIBS under the extreme condition of laser beam grazing the surface of non-flat samples. Examining the time-integrated emission spectra shows that enhancement factors of 2.5 and 3.5 were achieved at the laser energy of 33 mJ for aluminium alloy and brass, respectively. This factor first increases, reaches its maximum at 33 mJ, and drops gradually with the laser energy further increased. The mechanisms behind the enhanced optical emission and the enhancement factor evolution are discussed by using the proposed laser ablation model and laser-supported detonation (LSD) wave model, respectively. Examining the time-resolved emission spectra show that enhancement effect exists across all the plasma expansion process and the split beam mode allows for a longer plasma lifetime. A remarkable feature is that the trailing phenomenon emerging in the single beam mode vanishes due to the interaction between the plasmas generated by the grazing incident and normally incident laser beams in the split beam mode. The underlying cause is probably that the plasma plume produced by the normally incident laser beam prevents the grazing incident laser beam from further propagating and ablating the sample surface below. These findings not only give an insight into the plasma generation and evolution at grazing incidence of laser beam on sample surface but also provide a more reliable method for outdoor LIBS measurement of irregular samples.
© 2019 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Laser-induced breakdown spectroscopy (LIBS), as a new type of optical emission spectroscopy (OES) made possible by the invention of the laser, has become a well-established and versatile analytical technique during the past decades. The method is relying on focusing down the laser pulses to a sample surface so as to form high temperature plasma, and a portion of its emission is then collected to analyze the sample elemental composition in virtually any kind of sample (solid, liquid, gaseous, or aerosol) [1–5]. Benefited from its typical features such as simplicity, quasi non-destructive, simultaneous multi-element detection, absence of sample preparation, rapid in situ analysis and micro-analysis character of the measurements, LIBS has developed into a powerful analytical technique capable of providing qualitative and quantitative analysis in various fields, such as planetary exploration [2], industrial detection, geological analysis, agricultural, marine, and environmental science [6]. For example, LIBS technique has been successfully used in planetary exploration of Mars by the NASA rover Curiosity to seek life there [7]. Additionally, as for the elemental detection of nuclear materials and their isotopic concentrations, LIBS is in urgent need for monitoring nuclear waste and nuclear fuel production in environmental science [8].
In spite of the increasing popularity of traditional (single-pulse) LIBS, one of the issues that plague the LIBS technique is poor detection sensitivity which significantly limits the ability to make highly accurate quantitative measurements [9,10]. To overcome the limitation of single-pulse LIBS, many strategies were proposed to enhance the intensity of plasma emission, e. g., dual-pulse or ultrashort-pulse excitation, spatial and/or magnetic confinement, spark discharge, introduction of chamber, and ambient gas [11–15]. Among these techniques, dual-pulse scheme is most widely used because it provides a significant increase in the signal intensity of LIBS. However, using more than one laser source greatly increases the cost and size of the instrument. To avoid high cost and large size, Anotony et al. proposed a novel method of dual-wavelength LIBS technique by using a single laser system. They obtained 3 times enhancement of emission intensity for many emission lines of the lunar simulant samples [16]. In addition, Harilal et al. developed a collinear fs-DP-LIBS technique by dividing a laser beam into two branches in a Michelson interferometer system and they observed a significant signal enhancement of 3–4 times for Cu I and Zn I lines. From the economic point of view, the inexpensive ns-laser is preferable. Yang et al. presented a new LIBS technique based on a single-beam-splitting and obtained an enhancement factor of 5.6 and 4.8 for Al and Cu atomic emission lines respectively [17]. However, this signal enhancement could only be realized under the condition of laser energy above 60 mJ. When the laser energy is less than 60 mJ, the signal intensity of emission spectra obtained by single-beam-splitting didn’t increase but decrease, compared to the conventional single-beam LIBS measurement, which implies the unsuitability of single-beam-splitting scheme for signal enhancement using lower laser pulse energy.
In practical operations, LIBS is largely dominated by ambient conditions, of which the sample surface plays a critical role in controlling the sensitivity and reliability especially for trace elemental diagnosis and analysis. For the outdoor LIBS measurement, irregular targets, i.e., non-flat samples, are almost always observed, whose status is complex and unpredictable. Since the surface of samples is not flat as the preprocessed or standard samples in the laboratory, normal incidence cannot be ensured when the laser is focused on the sample surface. An incidence with larger angle will lead to lower laser ablation efficiency in LIBS measurement process since the decreasing laser irradiance causes a reduction in ablated mass, which largely weakens the emission intensity of laser-ablated plasma. This badly negative effects on LIBS measurement have been proved by several groups [18,19]. To reduce the influence of non-flat samples on LIBS measurement, the laser irradiance deposited on the local surface of samples needs to be as high as possible with the laser energy fixed at a certain value. Up to date, almost no method has been proposed to lower the influence of sample surface on LIBS and permit a more sensitive and more reliable LIBS technology .in this extreme condition of laser beam grazing the surface of non-flat samples. To fill this gap, a low-cost and convenient single-beam-splitting approach is employed to enhance the signal intensity of LIBS for non-flat samples by using a 1064-nm beam of Q-switched Nd:YAG laser. In contrast to the single-beam-split approaches mentioned above, ns-laser energy used in our experiments is no more than 60 mJ, at which a significant enhancement of emission intensity is achieved. Moreover, a new laser ablation model is proposed to explain the signal enhancement. Time-integrated emission spectra, time-resolved emission spectra, plasma parameters, and fast imaging of plasma plume are systemically investigated at various laser energies under different shooting modes.
2. Experimental methods
The schematic diagram of the experimental setup is shown in Fig. 1(a). The light source used for plasma generation is a 1064-nm beam of Q-switched Nd:YAG laser (Quantel, Ultra 100) with a repetition rate of 1–20 Hz and a pulse width of 8 ns full width at half maximum (FWHM). In this work, the repetition rate of laser was set at 20 Hz. The laser energy was monitored by measuring a small fraction of the incident laser beam (divided by BS 1, THORLABS BST11) with a laser energy meter (Nova II) due to the negligible diffraction effect. The laser energy used here varies from 18 to 48 mJ with the step of 5 mJ. Then the single laser pulse was divided into two beams by a beam-splitter (BS 2, THORLABS BST11) with an incident angle of 45°. After being reflected by mirror 1 and mirror 2, the reflection vertical laser beam I (70% of the total energy) is tangentially focused onto the sample surface via a 150 mm focal length lens 1 (focal spot diameter: 75 μm). The transmission horizontal laser beam II (30% of the total) was normally focused on the sample surface by another 150 mm focal length lens 2. With the optical path difference (OPD) between beam I and beam II equaling about 50 cm, the interval of the two beams approaching the sample is around 1.5 ns and less than the pulse width of laser (8 ns). Thus, it is reasonably assumed that the laser-induced plasmas are simultaneously generated by the two beams. Here, cylindrical aluminum alloy and cylindrical brass are selected to simulate the non-flat samples outdoors. The samples are mounted on a three-dimensional translation platform to permit the coincidence of the two laser beam focal points on the sample surface.
The laser-induced plasma produces optical emissions. When the emission spectra of the plasma are examined, the optical emissions are collected by an optical fiber (100 μm core diameter) fixed on a stage at an angle of 45° to both beam I and beam II, and coupled to a spectrometer (Andor Tech., Shamrock 303i) equipped with an intensified charge-coupled device (ICCD) camera (Andor Tech., iStar 340T, 512 x 1024 pixel). The spectrum range is set as required. For all the spectrum detection ranges, a grating of 2400 lines/nm was used to provide a spectral resolution of 0.05 nm. To observe the dynamic behavior of the plasma, the optical fiber is removed because of the space limitations and instead the mirror 3 is set in the same orientation to reflect the plasma optical emissions to the ICCD camera equipped with Nikon micro-lens. The front views of plasma images at different moments are obtained. When the ICCD camera is set at the place where the Nikon micro-lens is coaxial with the samples, the side view of plasma images can be acquired without using mirror 3. For both the spectrum and image examinations, the ICCD camera was operated in the gate mode. The gate delay and the gate width were adjusted to higher signal-to-background (SB) and signal-to-noise (SN) ratios for the time-integrated and time-resolved emission spectra of laser-induced plasmas. Prior to the experiment, the spectrometer calibration of wavelength was carried out with a mercury argon lamp. In addition, to verify the spectral correctness of LIBS, the wavelengths of the all spectral lines obtained were compared with NIST data. In order to reduce the standard deviation, an emission spectrum was obtained by averaging six sets of spectral data recorded for each condition.
3. Results and discussion
3.1. Enhancement of time-integrated emission spectra
The time-integrated emission spectra of laser ablated Al and Cu plasmas from metal samples were measured to demonstrate the enhancement performance in spectral intensity, where the delay time of 300 ns and the gate width of 100 μs were used, respectively. The plasma was produced by using four different shooting modes, i.e., beam I (solid curve, 23.1 mJ), beam II (short dashed curve, 9.9 mJ), split beam (beam I + beam II, dotted curve, 33 mJ), and single beam (blue circles, 33 mJ) with the same incident direction as that of beam I and the same energy as the sum of beam I + beam II, which are illustrated in Fig. 2. Figure 2(a) shows the emission spectra for the aluminium alloy in the range of 380–415 nm. Figures 2(b) and 2(c) illustrate the emission spectra for the brass in the ranges of 320–340 nm and 461–528 nm, respectively. Al I lines (394.40 nm and 396.15 nm), Cu I lines (324.75 nm, 327.39 nm, 465.11 nm, 510.55 nm, 515.32 nm, and 521.82 nm), and Zn I lines (330.26 nm, 472.21 nm, and 481.05 nm) were selected for spectral analysis due to their strong emission intensity. As seen in Fig. 2, all the emission spectral lines are enhanced in the split beam mode and different enhancement effects are observed for these lines. It is found from Fig. 2(a) that the spectral intensity of Al I lines is rather weak when a 23 mJ beam I is used to bombards the sample surface from the vertical direction (beam I mode). Increasing the energy of beam I to 33 mJ only leads to the enhancement of spectral intensity by nearly 2 times (single beam mode). But, splitting a 33 mJ beam into two orthogonal branches (beam I and beam II) and using the lower energy laser beam II to normally bombard the sample surface at the same time from the horizontal direction make the intensity enhanced more than 3.5 times (split beam mode). It is also found that the beam splitting yields about 2.5-fold signal enhancement when comparing with the single mode under the same laser energy of 33 mJ. As for the sample brass, the whole optical emission intensity is weakened to a great extent for all the shooting modes, but a better signal enhancement is obtained at these characteristic spectral lines (Cu I lines and Zn I lines) by using the split beam mode, as shown in Figs. 2(b) and 2(c). Compared to the beam I mode, an enhancement factor of 5 is achieved by adding another laser beam II with a small energy in the split beam mode. When comparing with the single mode, the enhancement factor can exceed 3.5 in the split beam mode. It should be noted that the beam I mode turns to be the single beam mode with increasing the energy of beam I to the value of 33 mJ, where beam II is removed. This energy value is the same as that of split beam mode.
The optical emission intensity is closely associated with the laser ablation efficiency that is largely determined by the laser irradiance. The laser irradiance is estimated according to the following formula:
where means the laser beam energy, is the FWHM of laser pulse and is the contact area between the sample surface and the transmission laser beam. A laser ablation model can be established to demonstrate the laser ablation efficiency under different shooting modes.As depicted in Figs. 1(b) and 1(c), means the contact area between the sample surface and beam I, while indicates the contact area between the sample surface and beam II. The area values are calculated without consideration of the optical system error and the sample surface smoothness. Since the focal spot of laser beam is very small, the contact area on the curved surface of the cylindrical sample is assumed to be a flat plane. Thus, and can be given as follow
Here, and denote the minor axis and major axis of the elliptical contact area, respectively. is the focal spot radius. is the angle between the vertical laser beam and the sample surface, which is assumed to be in our case. The laser irradiances in the split beam mode and in single beam mode are expressed aswhere and denote the laser energies of beam I and beam II, respectively. is the sum of and . For the beam I mode, the laser irradiance is the component of the laser irradiance, i.e., . In this work, takes the value of 8 ns. According to Eqs. (2) and (3), and are respectively estimated to be and at the laser energy of 33 mJ. Since is nearly four times the laser irradiance , the emission intensity enhancement in the split beam mode is ascribed to the increased laser irradiance and improved laser ablation efficiency. This is because that the ablated mass and the number of ablated particles are enlarged by growing heating of laser pulses leading to an increase in the intensity of OES [20].To verify the enhancement performance of spectral intensity under different pulse laser energies, the enhancement factor was examined with increasing the sum of the laser energy of beam I and beam II from 18 to 48 mJ. Figures 3 (a) and 3 (b) show the results for the samples aluminium alloy and brass, respectively. Here, four emission lines (Al I 394.40 nm, Al I 396.15 nm, Cu I 324.75 nm, and Cu I 327.39 nm) were selected as characteristic spectral lines. The enhancement factor means either the ratio of spectral intensity in the split beam mode to that in the beam I mode (: red & blue dashed lines) or the ratio of spectral intensity in the split beam mode to that in the single beam mode (: red & blue solid lines). It can be seen from Fig. 3 that the enhancement factor is highly dependent on the laser energy. As for the aluminium alloy, it follows from Fig. 3(a) that the enhancement factor first increases till reaching its maximum at the laser energy of 33 mJ, and then drops gradually with further increasing the energy. The average enhancement factor is 2.0 within the laser energy scope of interest. The rising tendency is possibly ascribed to more ablated mass per unit area or higher ablation efficiency due to the horizontal beam II. The decrease in the enhancement factor can be explained by the plasma formation model on solid surface, as illustrated in Fig. 4 [21]. When the incident irradiance is not much larger than the breakdown threshold, the LIBS plasma is typically in the laser-supported detonation (LSD) wave regime when expanding into atmospheric pressure air [22]. In this LSD wave model, the shock front is strong enough to heat the gas leading to absorbing the laser beam as the input of laser energy is increased. Figure 4(a) shows the plasma plume images of Al and Cu at a typical delay time of 300 ns in both the split beam mode and the single beam mode, which provides a reference for the laser ablation model about evolution of plasma under each condition, as shown in Figs. 4(b) and 4(c). It has been reported that the thickness of absorption zone decreases with the increasing angle of incident laser beam [23]. In the single beam mode, a large angle of incident laser beam allows for a thinness of absorption zone. The weak absorption effect ensures most of the vertical beam reaches the sample surface. In the split beam mode, one beam is split into two branches. As the laser energy rises, the plasma density is increased and the plasma turns to be opaque to the laser beam resulting in that parts of the horizontal beam II is blocked by a thicker absorption zone formed in the plasma. Thus, the increase of emission intensity with the laser energy is decelerated due to the absorption effect, which is responsible for the downward trend of enhancement factor with further increasing the laser energy.
But for the , the factor has a higher value (more than 4) at a lower laser energy because the ablation efficiency is quite small due to the grazing incidence of beam I on the sample surface. With increasing the laser energy from 18 to 43 mJ, a competition in ablating mass happens between beam I and beam II, which results in a fluctuation of enhancement factor. A further increase of laser energy makes the enhancement factor reduced because of the shielding effect on beam II. The average enhancement factor is 3.3. With respect to the sample brass, Fig. 3(b) shows a similar tendency of enhancement factor with increasing the laser energy. In contrast to the case for the sample aluminium alloy, the enhancement factors of Cu I lines have a higher average value, i.e., average enhancement factor of 2.5 and average enhancement factor of 4.5.
3.2. Enhancement of time-resolved emission spectra
Examining and comparing the time-integrated emission spectra obtained in four different shooting modes clearly demonstrate the signal enhancement performance of split beam mode in a period of time. To go further into the enhancement effect of this scheme at a specific moment, time-resolved emission spectra were investigated by measuring the optical emission spectra at distinct time intervals as the plasma decays and cools. Figures 5(a) and 5(b) respectively show the temporal evolutions of the emission intensities of Al I lines (394.40 nm and 396.15 nm) from aluminium alloy and Cu I lines (324.75 nm and 327.39 nm) from brass under two shooting modes, i.e., split beam (red & blue dashed lines) and single beam (red & blue solid lines). Here, the laser energy was set at 33 mJ and the emission intensities were obtained with the gate width of 300 ns and the increasing delay time from 0 to 5800 ns with a step of 300 ns. It should be noted that all the peak intensities were corrected by subtracting the background. As can be seen from Fig. 5(a), the emission intensity of Al I lines increases significantly in the initial stage of plasma expansion, reaches its maximum around 300 ns, and then decays gradually due to the cooling down of plasma in the following delay time. It is also found that the emission intensity in the split beam mode is much stronger than that in the single beam mode during all the delay time. This means that the enhancement of emission intensity due to the beam splitting exists across all the plasma expansion process. This enhancement feature is consistent with the observations in [24], where the combined spatial and magnetic confinement scheme was used to enhance the emission intensity. The optical emission lasts for 3000 ns in the single beam mode, but 5000 ns in the split beam mode, indicating that the beam splitting allows for a longer plasma lifetime. Similar or same lifetime elongation has been observed by using the magnetic confinement scheme [14] and the combined techniques of dual-pulse and magnetic confinement [25].
For the case of sample brass as shown in Fig. 5(b), the emission intensity is reduced when comparing with the case of sample aluminium alloy in the same shooting mode, which behaves as the same as that observed in the time-integrated emission spectrum examination. But a similar temporal evolution of emission intensities of Cu I lines and a similar enhancement performance of emission intensity are also observed. In addition, the Cu plasma lifetime is extended from 2500 to 4500 ns in the split beam mode.
3.3. Plasma parameters
The plasma temperature () and electron density () are from key plasma parameters that can be used to well understand the different excitation and ionization processes in plasma [26]. A point worth emphasizing is that and mentioned here should be regarded as indicatives of the average conditions occurring in the plume instead of the particular stage of its evolution since their values were determined by spectroscopic method. In this experiment, it is assumed that the Cu plasma is in local thermodynamic equilibrium (LTE), and then the plasma temperature can be deduced from Boltzmann plot method, which is given by the following expression [27]
where m and n denote the upper and lower levels for transition, respectively, is the wavelength, is the emission intensity, is the transition probability, is the statistical weight of upper level of the transition, and are the Plank constant and the speed of light in vacuum, is the energy of upper level, is the Boltzmann constant, is the total number density of species in plasma, and is the partition function. By plotting of Eq. (4) versus the energy of the upper level , the plasma temperature is calculated from the slope of the straight line () and can be evaluated without or .A series of Cu I lines, including Cu I (–) at 465.112 nm, Cu I (–) at 510.554 nm, Cu I (–) at 515.324 nm, and Cu I (–) at 521.820 nm, were selected to calculate the plasma temperature. All the relevant spectroscopic parameters of the transition lines are taken from the NIST database and summarized in Table 1 [28].
Figures 6(a) and 6(b) show the Boltzmann plots for the plasma temperature with the laser energy fixed at 33 mJ in the single beam mode and split beam mode, respectively. The plasma temperature was calculated to be 9400 K for the single beam mode and 9900 K for the split beam mode. It is found that the shooting mode has little influence on the plasma temperature. In addition, the same results were obtained with the laser energy set at 18 mJ and 48 mJ, as depicted in Fig. 6(c). The plasma temperature was estimated to be under all the conditions within the measurement uncertainty of 10%, which mainly comes from the uncertainties in the transition probabilities and the measurement of the spectral intensities of characteristic lines in the Boltzmann plots [29,30]. Comparing the plasma temperatures indicates that the variation in both the laser energy and the shooting mode produces little change in the plasma temperature. A similar behavior was also observed by other groups [31].
Based on the laser-supported radiation wave model, the plasma temperature can be deduced according to the equation below [32]:
, , and is the laser irradiance, the Stefan-Boltzmann constant, and the emissivity of the plasma, respectively. is the specific internal energy of plasma at the ignition point. denotes the correlation coefficient between the internal energy of the fully established plasma and the plasma temperature. In this equation, the ratio depends on the ionization potentials. In addition, the emissivity is related to the initial atom density. The plasma temperature depends weakly on the laser irradiance, atom density, and ionization potentials of components because the exponent in Eq. (5) is very small, which probably accounts for the little influence of both the laser energy and the shooting mode on the plasma temperature.
The electron density was determined by the Stark broadening measurement of the emission spectral lines. There are two main line broadening mechanics: Doppler broadening and Stark broadening. The Doppler broadening with Gaussian profile is mainly caused by electron impact and highly depends on the plasma temperature. The Doppler effect causes a variation of 0.01 nm in line width, which is too small to be considered here [20]. The Stark broadening with Lorentzian profile occurs when an emitting atom at a distance from an ion or electron is perturbed by the electric field [29]. As a result, the FWHM of the Stark broadened profile is related with the electron density. Here, we fit the emission spectral lines to the Voigt function and extract the FWHM of Lorentzian profile from the measured line width by deconvolution procedure. Then, the true width of the Stark broadening line is obtained by subtracting the instrumental line width from Lorentzian profile. Thus, the electron density () can be calculated through the relations [33]
The half width of the reduced Stark profile , which is related with electron density and plasma temperature, takes the value of [34]. The FWHM of the Lorentzian profile from instrumental broadening is measured to be 0.048 nm by utilizing a standard low pressure Hg calibration lamp [20]. Errors are bound to be present in determination of the electron density, leading to 15% uncertainty mainly ascribed to the uncertainties in the width measurement, the width deconvolution, and the related parameter [16].
Due to the outstanding features of strong, well-isolated, and free from self-absorption, Cu I line (–) at 324.754 nm was chosen for electron density determination [35]. As shown in Figs. 7(a) and 7(b), the fitted Voigt curve (red solid line) to the measured spectral line (blue circle) of Cu I under single beam mode and split beam mode was analyzed at the same laser energy of 33 mJ. In this case, the electron density was estimated to be , while, , and equal 0.372, 0.371, and 0.323 nm, respectively. By using the parallel treatment, the electron density in the split beam mode was obtained to be . This value has no apparent change comparing with that in the single beam mode, when the 15% uncertainty was taken into account. These results suggest that the electron density has weak dependence on the shooting mode. The electron density was also examined under the laser energy of 18 mJ and 48 mJ, as depicted in Fig. 7(c). Although the electron density increases due to the growing heating of laser pulses and the increasing ablated mass with laser energy, it changes little between the two shooting modes.
As mentioned above, the measurement of plasma temperature is based on the assumption of local thermodynamic equilibrium (LTE). This hypothesis requires sufficient electron density to ensure the high collision rate so that the collisional process dominates the condition of the atomic and ionic states. The lower limit of electron density is given by McWhirter criterion [33]:
where is the energy difference between the two states. The maximum is 2.67 eV for Cu I at 465.112 nm and is about 9500 K. The low limit of electron density was calculated to be , which is much smaller than that () obtained in our case. This indicates that our experiment condition fulfills the LTE.3.4. Fast imaging of laser-induced Al and Cu plasmas
In order to evaluate the effect of different shooting modes on the physical profiles of the generated plumes, the time-resolved photographic images of the plasma plume from aluminium alloy and brass were directly acquired using the ICCD camera, with the results comparatively shown in Fig. 8. It is noteworthy that all the images were handled by a false-color technique encoding the emission intensity normalized to the maximum intensity of image and each image made with a separate laser pulse [36].The images were taken at various delay times with different gate widths. Before the delay time 200 ns, the images were obtained with a gate width of 10 ns and a step of 20 ns. After the delay time 200 ns, the gate width and step were set at 50 ns and 100 ns, respectively. In the split beam mode, the laser energies of beam I and beam II were set to 23.1 mJ and 9.9 mJ, respectively. As for the single beam scheme, beam I was fixed at 33 mJ, To understand the dynamic behavior of laser plasma more clear, the incident laser beams (beam I and/or beam II) are labeled around the expanded plasma plume for the two shooting modes. As can be seen from these images, for both the two modes, the plasma plume gradually becomes brighter and brighter from 10 to 300 ns. With increasing the delay time to 1000 ns, although the plasma plume expands outwards and the volume increases, its brightness begins to decrease due to the cooling down of plasma. During the whole process, the plasma plume is brighter and larger in the split beam mode [37], and this feature is especially obvious for the delay time after 160 ns as depicted in Figs. 8(a) and 8(b). The dynamic behavior of plasma plume is consistent with the observations of time-resolved emission spectra, where the spectral intensity reaches its maximum around 300 ns and is enhanced in the split beam mode. The larger plasma plume also illustrates a faster expansion process, which accelerates the reduction of the electron density. This is probably responsible for the fact that at a certain laser energy, the electron density nearly has the same value in the two modes.
With respect to the shape of plasma plume, it is found from Fig. 8(a) that the evolving plasma plume is approximately ellipsoidal in profile and its axis is along the opposite direction of the incident laser beam (beam I), which was also reported by other groups [21,36,38]. Since the incident beam (beam I) almost grazes the surface of cylindrical sample, laser beam with a considerable part of energy passages through the focus point to interact with the sample surface below. Thus, the plasma plume exhibits an obvious “tail” below the core of plume in the single beam mode. This tail gradually appears with the delay time increased from 20 to 300 ns, and then it is separated from the main body at the delay time of 400 ns. This feature lasts for 200 ns. Afterwards, this trailing phenomenon disappears and the plasma diffuses, leading to decrease in the brightness till the delay time of 1000 ns. But for the split beam mode, the “tail” cannot be observed. This is probably attributed to fact that the plasma plume produced by beam II prevents beam I from further propagating forward and ablating the sample surface below.
Figure 8(b) shows the case for the sample brass. Compared to that of aluminium alloy, a similar result about the temporal evolution of plasma plume is observed in both the two shooting modes. However, as regards to the size of plume, plasma from brass sample is much smaller than Al plasma plume at the same delay time and gate width, which is in good agreement with the time-integrated OES obtained above: The spectral intensities of Al I lines are several times greater than those of Cu I lines with the same time delay and gate width.
4. Conclusions
Summarily, a single-beam-splitting approach was employed to enhance the signal intensity of LIBS under the extreme condition of laser beam grazing the surface of non-flat samples. Examining the time-integrated emission spectra of aluminium alloy and brass samples show that significant enhancement in the spectral intensity of Al and Cu plasmas were observed by using the single-beam-splitting approach. Compared to the single mode at the same laser energy of 33 mJ, the split mode makes the spectral intensity improved by 2.5 and 3.5 times for the aluminium alloy and brass samples, respectively. This is attributed to the higher ablation efficiency under a higher laser irradiance leading to a higher concentration of the excited atoms in the split beam mode. The enhancement factors of spectral intensity under different shooting modes, and , were also examined with the laser energy increased from 18 to 48 mJ. It is found that the signal enhancement effect is highly dependent on the laser energy. For both the Al I lines and Cu I lines, the enhancement factors first increase, then reach their maxima at the laser energy of 33 mJ, and drop gradually with the energy further increased. The rising tendency of the enhancement factor is ascribed to more ablated mass per unit area or higher ablation efficiency in the split mode, while the decrease in the enhancement factor is resulted from the increasing plasma shielding effect on the sample surface and a stronger absorption of beam II energy. For , the enhancement factor presents a higher value at a lower laser energy and fluctuates with further increasing the laser energy, which respectively results from the small ablation efficiency under the condition of the grazing incidence of beam I on the sample surface and the competition in ablating mass between beam I and beam II. Examination of the time-resolved emission spectra of the samples indicates that the enhancement of emission intensity in the split beam mode exists across all the plasma expansion process and beam splitting allows for a longer plasma lifetime. It is also found that the shooting mode has little influence on both the plasma temperature and electron density, which are respectively due to the weak dependence of plasma temperature on the laser irradiance and the faster plasma expansion process depressing the particle density in the split beam mode.
The fast images of the plasma plumes show that for both the single beam and split beam modes, the plasma plume expands outwards and the volume increases. Its brightness increases from 10 to 300 ns, but decreases due to the cooling down of plasma with further increasing the delay time to 1000 ns. In contrast to the single beam mode, the split beam mode allows for a brighter and larger plasma plume during the whole process. Moreover, the trailing phenomenon emerging in the single beam mode is eliminated by using beam splitting, which is probably because that the plasma plume produced by beam II prevents beam I from further propagating forward and ablating the sample surface below.
Our work suggests that a significant enhancement of signal intensity can be achieved by using the single-beam-splitting approach in LIBS analysis of irregular samples under the condition of laser energy less than 60 mJ. This approach lowers the influence of sample surface on the LIBS measurement and improves the LIBS analytical sensitivity to some extent under extreme conditions. Thus, a more reliable scheme is provided for the qualitative and quantitative analysis of outdoor LIBS with the advantage of low cost.
Funding
National Natural Science Foundation of China (51877210, 61475191, 61875228); Chinese Academy of Sciences (CAS) “Light of West China” Program (XAB2015A08); the Open Research Fund of Key Laboratory of Spectral Imaging Technology, CAS (LSIT201807G).
References
1. T. Xu, Y. Zhang, M. Zhang, Y. He, Q. Yu, and Y. Duan, “Temporal-resolved characterization of laser-induced plasma for spectrochemical analysis of gas shales,” Spectrochim. Acta B At. Spectrosc. 121(1), 28–37 (2016). [CrossRef]
2. J. Bergevin, T. H. Wu, J. Yeak, B. E. Brumfield, S. S. Harilal, M. C. Phillips, and R. J. Jones, “Dual-comb spectroscopy of laser-induced plasmas,” Nat. Commun. 9(1), 1273 (2018). [CrossRef] [PubMed]
3. Y. Tian, B. Xue, J. Song, Y. Lu, and R. Zheng, “Non-gated laser-induced breakdown spectroscopy in bulk water by position-selective detection,” Appl. Phys. Lett. 107(11), 111107 (2015). [CrossRef]
4. J. Guo, Y. Lu, K. Cheng, J. Song, W. Ye, N. Li, and R. Zheng, “Development of a compact underwater laser-induced breakdown spectroscopy (LIBS) system and preliminary results in sea trials,” Appl. Opt. 56(29), 8196–8200 (2017). [CrossRef] [PubMed]
5. S. S. Harilal, P. K. Diwakar, and A. Hassanein, “Electron-ion relaxation time dependent signal enhancement in ultrafast double-pulse laser-induced breakdown spectroscopy,” Appl. Phys. Lett. 103(4), 041102 (2013). [CrossRef]
6. Y. Li, Y. Lu, Y. Lan, Y. Li, J. Guo, and R. Zheng, “Sr/Ca ratio analysis of seashells using laser-induced breakdown spectroscopy under objective-lens focusing and single-lens focusing,” Appl. Opt. 57(13), 3539–3545 (2018). [CrossRef] [PubMed]
7. R. C. Wiens, S. Maurice, B. Barraclough, M. Saccoccio, W. C. Barkley, J. F. Bell, S. Bender, J. Bernardin, D. Blaney, J. Blank, M. Bouyé, N. Bridges, N. Bultman, P. Caïs, R. C. Clanton, B. Clark, S. Clegg, A. Cousin, D. Cremers, A. Cros, L. DeFlores, D. Delapp, R. Dingler, C. D’Uston, M. Darby Dyar, T. Elliott, D. Enemark, C. Fabre, M. Flores, O. Forni, O. Gasnault, T. Hale, C. Hays, K. Herkenhoff, E. Kan, L. Kirkland, D. Kouach, D. Landis, Y. Langevin, N. Lanza, F. LaRocca, J. Lasue, J. Latino, D. Limonadi, C. Lindensmith, C. Little, N. Mangold, G. Manhes, P. Mauchien, C. McKay, E. Miller, J. Mooney, R. V. Morris, L. Morrison, T. Nelson, H. Newsom, A. Ollila, M. Ott, L. Pares, R. Perez, F. Poitrasson, C. Provost, J. W. Reiter, T. Roberts, F. Romero, V. Sautter, S. Salazar, J. J. Simmonds, R. Stiglich, S. Storms, N. Striebig, J.-J. Thocaven, T. Trujillo, M. Ulibarri, D. Vaniman, N. Warner, R. Waterbury, R. Whitaker, J. Witt, and B. Wong-Swanson, “The ChemCam instrument suite on the Mars Science Laboratory (MSL) Rover: Science objectives and mast unit description,” Space Sci. Rev. 170, 95–166 (2012). [CrossRef]
8. M. Z. Martin, S. Allman, D. J. Brice, R. C. Martin, and N. O. Andre, “Exploring laser-induced breakdown spectroscopy for nuclear materials analysis and insitu applications,” Spectrochimca Acta Part B. 74, 177–183 (2012). [CrossRef]
9. Z. Wang, T. B. Yuan, Z. Y. Hou, W. D. Zhou, J. D. Lu, H. B. Ding, and X. Y. Zeng, “Laser-induced breakdown spectroscopy in China,” Front. Phys. 9(4), 419–438 (2014). [CrossRef]
10. R. K. Singh and J. Narayan, “Pulsed-laser evaporation technique for deposition of thin films: Physics and theoretical model,” Phys. Rev. B Condens. Matter 41(13), 8843–8859 (1990). [CrossRef] [PubMed]
11. X. W. Li, Z. Wang, X. L. Mao, and R. E. Russo, “Spatially and temporally resolved spectral emission of laser-induced plasmas confined by cylindrical cavities,” J. Anal. At. Spectrom. 29(11), 2127–2135 (2014). [CrossRef]
12. V. Zorba, X. Mao, and R. E. Russo, “Femtosecond laser induced breakdown spectroscopy of Cu at the micron/sub-micron scale,” Spectrochimca Acta Part B. 113(1), 37–42 (2015). [CrossRef]
13. V. Zorba, J. Syzdek, X. Mao, R. E. Russo, and R. Kostecki, “Ultrafast laser induced breakdown spectroscopy of electrode/electrolyte interfaces,” Appl. Phys. Lett. 100(23), 234101 (2012). [CrossRef]
14. X. K. Shen, J. Sun, H. Ling, and Y. F. Lu, “Spatial confinement effects in laser-induced breakdown spectroscopy,” Appl. Phys. Lett. 91(8), 081501 (2007). [CrossRef]
15. A. Huassin, X. Gao, Z. Hao, and J. Lin, “Combined effects of double pulses and magnetic field on emission enhancement of laser-induced breakdown spectroscopy from aluminum plasma,” Optik (Stuttg.) 127(20), 10024–10030 (2016). [CrossRef]
16. J. K. Antony, N. J. Vasa, V. L. N. S. Raja, and A. S. Laxmiprasad, “Single laser based dual-wavelength ablation technique for emission enhancement during LIBS,” J. Phys. D Appl. Phys. 45(36), 365401 (2012). [CrossRef]
17. Y. Li, D. Tian, Y. Ding, G. Yang, K. Liu, C. Wang, and X. Han, “A review of laser-induced breakdown spectroscopy signal enhancement,” Appl. Spectrosc. Rev. 53(1), 1–35 (2017). [CrossRef]
18. F. Colao, R. Fantoni, V. Lazic, A. Paolini, F. Fabbri, G. G. Ori, L. Marinangeli, and A. Baliva, “Investigation of LIBS feasibility for in situ planetary exploration: An analysis on Martian rock analogues,” Planet. Space Sci. 52(1), 117–123 (2004). [CrossRef]
19. G. Nicolas, M. P. Mateo, and V. Pinon, “3D chemical maps of non-flat surfaces by laser-induced breakdown spectroscopy,” J. Anal. At. Spectrom. 22(10), 1244–1249 (2007). [CrossRef]
20. B. Lei, J. Wang, J. Li, J. Tang, Y. Wang, W. Zhao, and Y. Duan, “Influence of humidity on the characteristics of laser-induced air plasma,” Jpn. J. Appl. Phys. 57(10), 106001 (2018). [CrossRef]
21. D. A. Cremers and R. A. Multari, Handbook of Laser-Induced Breakdown Spectroscopy (JohnWiley & Sons Ltd, 2006).
22. S. Musazzi and U. Perini, Laser-Induced Breakdown Spectroscopy (Springer, 2014)
23. G. Yang, Q. Lin, Y. Ding, D. Tian, and Y. Duan, “Laser induced breakdown spectroscopy based on single beam splitting and geometric configuration for effective signal enhancement,” Sci. Rep. 5(1), 7625 (2015). [CrossRef] [PubMed]
24. L. B. Guo, W. Hu, B. Y. Zhang, X. N. He, C. M. Li, Y. S. Zhou, Z. X. Cai, X. Y. Zeng, and Y. F. Lu, “Enhancement of optical emission from laser-induced plasmas by combined spatial and magnetic confinement,” Opt. Express 19(15), 14067–14075 (2011). [CrossRef] [PubMed]
25. D. Wu, P. Liu, L. Sun, R. Hai, and H. Ding, “Influence of a static magnetic field on laser induced tungsten plasma in air,” Plasma Sci. Technol. 18(4), 364–369 (2016). [CrossRef]
26. S. S. Harilal, C. V. Bindhu, R. C. Issac, V. P. N. Nampoori, and C. P. G. Vallabhan, “Electron density and temperature measurements in a laser produced carbon plasma,” J. Appl. Phys. 82(5), 2140–2146 (1997). [CrossRef]
27. W. Luo, W. Zhao, Y. Duan, and H. Wang, “Diagnostics of air plasma ablated by 1064 nm laser pulses,” Chin. Opt. Lett. 9(s1), s10303 (2011). [CrossRef]
28. National Institute of Standards and Technology, https://physics.nist.gov/cgi-bin/ASD/energy1.pl.
29. M. Hanif, M. Salik, and M. A. Baig, “Quantitative studies of copper plasma using laser induced breakdown spectroscopy,” Opt. Lasers Eng. 49(12), 1456–1461 (2011). [CrossRef]
30. N. M. Shaikh, S. Hafeez, M. A. Kalyar, R. Ali, and M. A. Baig, “Spectroscopic characterization of laser ablation brass plasma,” J. Appl. Phys. 104(10), 103108 (2008). [CrossRef]
31. N. Kawahara, J. I. Beduneau, T. Nakayama, E. Tomita, and Y. Ikeda, “Spatially, temporally, and spectrally resolved measurement of laser-induced plasma in air,” Appl. Phys. B 86(4), 605–614 (2007). [CrossRef]
32. S. Yalçin, D. R. Crosley, G. P. Smith, and G. W. Faris, “Influence of ambient conditions on the laser air spark,” Appl. Phys. B 68(1), 121–130 (1999). [CrossRef]
33. W. Luo, Q. Sun, C. Gao, J. Tang, H. Wang, and W. Zhao, “Plasma properties of 532 nm laser-ablated aluminum E414d target with different power densities,” Plasma Sci. Technol. 12(4), 385–390 (2010). [CrossRef]
34. H. R. Griem, Spectral Line Broadening by Plasma (Acdemic, 1964).
35. A. M. El Sherbini, H. Hegazy, and T. M. El Sherbini, “Measurement of electron density utilizing the Hα-line from laser produced plasma in air,” Spectrochim. Acta B At. Spectrosc. 61(5), 532–539 (2006). [CrossRef]
36. D. H. Zhang, X. X. Yuan, M. G. Su, Q. Min, C. Z. Dong, and D. X. Sun, “Shielding and diagnostics of laser-induced air plasmas generated in collinear double pulse configuration,” Phys. Plasmas 25(6), 063112 (2018). [CrossRef]
37. X. L. Mao, X. Z. Zeng, S. Wen, and R. E. Russo, “Time-resolved plasma properties for double pulsed laser-induced breakdown spectroscopy of silicon,” Spectrochim. Acta B At. Spectrosc. 60(7), 960–967 (2005). [CrossRef]
38. R. Hai, X. L. Mao, G. C. Y. Chan, R. E. Russo, H. B. Ding, and V. Zorba, “Internal mixing dynamics of Cu/Sn-Pb plasmas produced by femtosecond laser ablation,” Spectrochim. Acta B At. Spectrosc. 148, 92–98 (2018). [CrossRef]