2020 Volume 60 Issue 12 Pages 2845-2850
This paper suggests a method to control the focal point of laser on the on-focus position of a sample surface automatically in laser-induced breakdown spectrometry (LIBS). For this purpose, an electrically-tunable plano-convex lens was installed in a laser irradiation system, where it could vary the focal length of laser with a long working distance and a rapid response time, and the focal length could be periodically varied with a triangle waveform. Because the tunable lens was easily handled and inexpensive, the laser system could be modified with a low cost, as compared with commercial apparatuses having complicated optics to control the position of laser irradiation. A piece of scrapped stainless steel the surface of which was titled and had some roughness was investigated as a test specimen. A satisfactory result was obtained such that the plasma could be generated uniformly and firmly along a laser trace on the sample surface and thus could give the emission signal with a sufficient precision. The driving frequency of the tunable lens, which controlled a repetition period of the laser beam, was optimized to be 10 Hz when the scan rate of laser was fixed at 3.0 mm/s. As a result, it is expected that the LIBS system with the tunable lens can be applied to actual on-site/in-line analysis in material production.
A gaseous plasma generated by irradiating a high-power laser is extensively employed as an excitation source for optical emission spectrometry, generally called laser-induced breakdown spectrometry (LIBS).1) LIBS has several features suitable for on-site/in-line analysis in the actual process of material production, in which an as-received sample can be analyzed with little pre-treatment under ambient air atmosphere and thus the analytical result can be obtained easily and rapidly. In steelmaking industry, it is now expected that LIBS can be applied to the ladle analysis of steel products as an alternative to spark discharge plasma optical emission spectrometry.2) Also, LIBS is a promising analytical method for sorting scrapped metallic materials because it can determine their chemical compositions with a high response time,3) enabling the metal resources to be highly recycled on the basis of the analytical information.
It is an important issue in LIBS that the focal position of laser irradiation has to be strictly controlled to create laser-induced plasma on a sample surface; otherwise, no emission of the sample species can be observed due to insufficient focusing of the laser beam. Generally, the focal point of laser should be controlled with a precision of ca. 1 mm, dependent of the incident power, the wavelength and the oscillation mode of laser. This would cause a serious problem especially when LIBS is applied to in-line analysis, where lots of specimens having various sizes and shapes have to be recognized and analyzed in a short measuring time. For such a purpose, Noll et al. suggested an experimental apparatus of LIBS including an auto-tuning system of the laser focal position to detect the emitted radiation from the specimens.4) More recently, a group of Deguchi published a LIBS measuring system for online elemental monitoring of a steel manufacturing processes,5) and also reported a LIBS analysis of manganese in steel samples under high temperature atmosphere.6) A LIBS apparatus, which has a complicated optical system so that the emitted radiation as well as the laser beam can be simultaneously focused just on a sample surface, has been commercially available for the fast in-line analysis.7) This apparatus can provide analytical results enough to distinguish individual specimens while they are moving on a conveyer; however, high initial and running costs are needed to operate it.
We consider a simple and inexpensive system of LIBS by using a fast electrically tunable lens, which may be available for sorting scrapped materials when the specimens have been arranged in a line by any mechanical operation. Figure 1(a) indicates a simplified illustration of the electrically tunable lens. The tunable lens changes in shape by an external force. It consists of a container, which is filled with an optical fluid and sealed off with an elastic polymer membrane.8) The deflection of the lens is proportional to the pressure in the fluid, which is controlled by an electromagnetic actuator that is used to exert pressure on the container.8) As a result, the focal distance of the lens is varied by the current flowing through the coil of the actuator. When the lens is installed in a LIBS apparatus, the focal point of laser can be changed widely and rapidly along the incident direction of the laser beam. In this study, we represent the performance of this LIBS system, especially regarding auto-focusing of the laser irradiation when a sample moves along one direction (the laser is scanned on the sample surface).
Schematic drawing of the operation of an electrically tunable lens (a), the variation of the focal length driven by applied current (b), and a cyclic variation of the focus controlled by a timing pulse from a function generator (c). (Online version in color.)
A schematic diagram of the experimental setup is depicted in Fig. 2. A Q-switched pulsed Nd: YAG laser (SSL-330-50, EKSPLA, Lithuania) was operated at a pulse width of about 150 ps and an output wavelength of 1064 nm (fundamental frequency). A repetition rate of 50 Hz was applied with a laser power of about 20 mJ/pulse. The energy of the pulsed laser was measured using a power meter for lasers with a thermal sensor (3A-P, Ophir Photonics, Israel). The laser beam was focused on the sample surface with a fast electrically-tunable plano-convex lens (EL-10-30, Optotune, Switzerland), having an external diameter of 30 mm and a clear aperture of 10 mm. This lens transmits 95% of the laser beam at a wavelength of 1064 nm, and the adjustable range of the focal distance may be varied from lens to lens;8) therefore, we calibrated the actual working range of the lens employed in our experiment to obtain a variation from 20 to 8 mm when the current applied to the electromagnetic actuator increased from 0 up to 250 mA, as shown in Fig. 1(b). The adjustable range was also dependent on temperature. The response time in a pulsed operation is less than 2.5 ms when a rectangular current step changes the focal length from 10 to 90% as much as the maximum value,8) indicating that the lens can follow a cyclic variation in the applied current at a frequency of several 100 Hz (see Fig. 1(c)).
Block diagram of a LIBS apparatus employed in this study. (Online version in color.)
The plasma emission was focused on the entrance port of an optical fiber and collected on the entrance slit of a spectrometer, dispersed by a Czerny-Turner-type monochromator (MS7504i, SOL Instruments Ltd., Belarus) with a grating of 3600 lines/mm at an actual resolution of 0.02 nm, and detected by an intensified charge-coupled detector (ICCD) (DH734-18F-03, Andor, UK). All the intensity data were acquired during a gate width of 500 μs at a delay time of 0.5 μs after each shot of the pulsed laser, so that strong background emission just after the breakdown could be blocked.9) All the experiments were carried out under ambient air atmosphere and flow of the atmospheric gas was not controlled.
Eight standard reference materials (SRM) of aluminum alloy (Japan Aluminium Association, Tokyo) were prepared to investigate the characteristics of the focal-tunable lens. In this study, a major alloyed element of magnesium was principally determined in the SRMs having the magnesium content of 0.008, 0.014, 0.57, 0.97, 1.29, 1.38, 2.8, and 4.24 mass%. A sensitive atomic line, the Mg I 285.21-nm line, was selected as the analytical line. These disk-shaped SRMs had flat surfaces, and all the measurements were carried out for their as-received surfaces without any pre-treatments before LIBS, such as polishing with emery paper. In addition, a piece of scrapped stainless steel was prepared to evaluate auto-focusing of the laser irradiation. As a photograph is shown in Fig. 3, this specimen has a dimension of 50 × 40 × 30 mm, with the surface tilted at an angle of about 20 degrees. As similar to the SRMs of aluminum alloy, the as-received surface was under irradiation of the laser beam.
Photograph of a test specimen (a piece of scrapped stainless steel). (Online version in color.)
These specimens were set on a motor-driven X-Y-Z stage (SGSP26-200S, SIGMA Koki, Japan), which was controlled with a system controller (SHOT-304GS, SIGMA Koki, Japan), and could be precisely moved to the laser beam in three-dimensional directions. The X-Y-Z stage was first adjusted in the Z-direction (height direction) until the irradiation position of laser had been focused just on a target position of the sample surface, and then the tunable lens worked to vary the focal position of laser in a cyclic manner when the specimen was moved along a lateral direction at a constant speed. Our investigation was interested in how working parameters of the tunable lens should be optimized, in order that the tilted sample of stainless steel could provide the emission signal even when the surface position was varied.
First of all, we investigated a variation in the emission intensity of an Al I emission line in a SRM of aluminum alloy when the focal point of laser was step-wisely changed at an interval of 0.2 mm. The Al I 309.27-nm line was employed as the analytical line. Figure 4 shows a typical result, where the displacement of 0 mm indicates the on-focus position when the laser was focused just on the surface of specimen, and the plus and minus signs mean focal positions above and below the surface, respectively. It was interesting that the emission intensity became drastically reduced at focal positions just above the surface (only 2 mm apart from the surface), because the irradiation energy of laser would be largely absorbed by the surrounding gas, thus decreasing atoms of the specimen to be ablated and ejected from the surface. On the other hand, the emission intensity could be observed at focal positions beneath the surface (down to 6 mm from the surface). The reason for this effect is that LIBS plasma can be produced even at these off-focus positions, while the laser energy becomes insufficiently transferred to the specimen due to de-focusing of the laser beam. In these cases, electrons of the metallic sample might work as a trigger in a production of the LIBS plasma in which atoms of the specimen are involved.
Variation in the emission intensity of the Al I 309.27 nm in an aluminum alloy when the focus point of laser is gradually changed at an interval of 0.2 mm. A relative distance of 0 mm corresponds to the on-focus position on the sample surface.
It is found in Fig. 4 that the focal position should be regulated to be within ±1 mm from the surface of specimen (the on-focus point). A peak of the emission intensity was observed in this range of the focal position. The focus-tunable lens enables this condition to be automatically fulfilled by varying the focal distance of laser in a cyclic manner. However, experimental parameters of the lens, such as the frequency and the working distance of the focal point, strongly affect the resultant emission intensities, and thus they should be optimized.
3.2. Comparison in a Calibration Curve between Fixed and Cyclic-varying Focal PointsWe investigated a calibration relationship between the emission intensity of the Mg I 285.21 nm and the magnesium content in the aluminum alloy SRMs when the tunable lens worked in a cyclic manner, in order to evaluate the analytical performance in comparison to the result when the focal point was fixed on the on-focus position. As illustrated in Fig. 1(c), the tunable lens was regulated using a triangle wave which was generated with a function generator at a fixed frequency and a fixed amplitude; therefore, the focal point was varied linearly and repeatedly.
For a typical instance, the tunable lens was modulated at a frequency of 0.25 Hz and a working distance of ±10 mm from the on-focus position, where the laser breakdown occurred intermittently, and the emission intensity from the plasma was measured for 16 s. This condition for the lens realized that the laser irradiation crossed the on-focus position 8-times, and that about 80 laser shots could contribute to the plasma formation when the laser beam was irradiated at a repetition rate of 50 Hz, under the assumption that the effective range of the focal position was ±1.0 mm as denoted in Fig. 4. The emission intensity of the Mg I line was integrated and averaged over the measuring time of 16 s, and four replicate measurements were conducted for each aluminum alloy SRM. Figure 5(a) shows a calibration curve of the averaged intensities in the auto-focusing operation with the tunable lens using seven of the SRMs (the data of the 1.38-mass% SRM was excluded due to a measurement error), in which their relative standard deviations were estimated in a range between 4 to 18%. The emission intensity became saturated at higher contents of magnesium. This saturation is due to self-absorption of the Mg I 285.21-nm line, because this emission line is assigned to a resonance transition from 3p 1P1 (4.3457 eV) to 3s 1S0 (0.00 eV).10) However, the resultant calibration curve could be well fitted to a second-order polynomial with a correlation coefficient (r2) of 0.9713. For comparison, a calibration curve was estimated when 50 shots of the laser were fixed to the on-focus position of specimens, as shown in Fig. 5(b). As similar to the data of the auto-focusing laser (see Fig. 5(a)), a fitting with a second-order polynomial could be performed with a correlation coefficient (r2) of 0.9836. Therefore, the calibration relationship could be determined with high reliability in the auto-focusing mode as well as the fixed on-focus mode, implying that an auto-focusing control of laser could be applied to the quantitative analysis of LIBS.
Calibration curves for the emission intensity ratio of Mg I 285.21 nm versus the Mg content in seven SRMs of aluminum alloy, when the focal position of laser is modulated at a frequency of 0.25 Hz (a) and is fixed on the on-focus position of the sample surface (b). (Online version in color.)
A repetition rate of laser, a moving speed of the sample stage (a scan speed of the laser beam), and a driving frequency of the tunable lens are interrelated experimental parameters in the auto-focusing operation. More laser shots at a higher repetition rate are needed for observing a dynamic variation of the emission intensity along with the movement of the specimen. Therefore, the laser repetition rate was selected to be 50 Hz, since it was the maximum rate of our laser system. The other parameters were determined under various possible combinations. In this measurement, the moving speed of the stage was fixed at 3.0 mm/s, enabling 500 laser shots to hit on the specimen surface at a sampling distance of 30 mm. As seen in Fig. 3, the piece of scrapped stainless steel analyzed in this study had a tilted and distorted surface; therefore, the focal points of laser were always varied with moving the sample. The tunable lens works to adjust the focal distance of laser irradiation so that the laser beam can be focused on the specimen surface as frequently as possible; for this purpose, the driving frequency of the lens should be optimized.
A large number of emission lines, originated from iron, chromium, and nickel in the specimen of stainless steel, were found in the LIBS spectrum. Two sensitive lines of the Cr I 360.53 nm and Fe I 358.12 nm were simultaneously measured along with a scan of the laser beam on the specimen surface. The intensity ratio of the Cr I to the Fe I lines was estimated to correct an ablated amount of sample atoms for each laser shot. The driving frequency of the tunable lens was selected to be 0.25, 0.5, 1, 2, 5, 10, 25, or 50 Hz, with a working distance of ±10 mm from the on-focus position. It was measured how many laser shots could contribute to the plasma creation enough to excite the atomic emission of the sample atoms. Here, we defined ‘an effective shot’ to have the Cr I/Fe I intensity ratio more than a threshold value, which was the average value minus the standard deviation when they were calculated over all the laser shots (nominally 500). Figure 6 represents that a percentage of the effective shots to the whole laser shots is the largest at a driving frequency of 10 Hz, which is a desirable condition for the laser scan. The intensity ratio was integrated and averaged for each laser scan at a fixed distance of 30 mm on the specimen surface. Then, triplicate measurements were carried out for each driving frequency to check the precision and the repeatability. Figure 7 indicates plots of the integrated intensity ratio (a) and its relative standard deviation (RSD) (b) in the triplicate measurements, as a function of the driving frequency of the tunable lens. The emission intensity has a maximum value at a driving frequency of 10 Hz, and that the relative standard deviation is larger with an increase of the driving frequency, meaning that the data precision becomes worse at higher driving frequencies. The following discussion is described to understand the reason for these results.
Percentage of the effective shots resulting in plasma to the whole shots at several driving frequencies of tunable lens, where error bars were estimated from triplicate measurements. The laser of 50 Hz was scanned on the surface of a piece of stainless steel at a moving speed of 3.0 mm/s.
Plots of the integrated value of the intensity ratio of Cr I 360.53/Fe I 358.12 (a) and the relative standard deviation (b), as a function of the driving frequency of the tunable lens. The laser of 50 Hz was scanned on the surface of a piece of stainless steel at a moving speed of 3.0 mm/s.
At lower driving frequencies (0.25 and 0.5 Hz), the vertical motion of the laser beam is slow, which may decrease the probability that the laser beam can reach the on-focus positions while the specimen is moved. This is a possible reason why the averaged emission intensity as well as the number of the effective laser shot is reduced at the low driving frequency of the tunable lens. On the other hand, at higher driving frequencies (25 and 50 Hz), the vertical motion of the laser beam is so fast that the irradiated laser beam might not provide sufficient amounts of sample atoms (the laser ablation might occur faintly), thus resulting in the lower emission intensities. Figure 8(a) shows microscope images of the ablation groove left on the specimen surface at driving frequencies of 10 Hz and 50 Hz. The groove of 50 Hz is more continuous and thinner compared to that of 10 Hz. This observation is an evidence that, at the 50-Hz driving frequency, the laser beam could hit the on-focus positions frequently but the resultant plasma would be insufficient to ablate the sample atoms and then to emit the radiation. A model for the plasma production is represented in Fig. 8(b). At lower driving frequencies, the plasma breakdown occurs intermittently; however, each breakdown may produce the plasma more firmly such that the laser energy can be sufficiently transferred into the sample surface. This effect might explain the result regarding RSD as shown in Fig. 7(b): the emission intensity can be measured more precisely at the lower driving frequencies.
Optical microscope images (a) of the ablated groove when the laser beam is modulated at driving frequencies of 10 and 50 Hz. Schematic drawing (b) for representing cyclic variations of the laser beam at higher (left) and lower (right) driving frequencies of the tunable lens. (Online version in color.)
Our measurement recommends that the driving frequency of the tunable lens is set at 10 Hz, because the emission intensity from the specimen is the most intense with a relatively low RSD (2.5%). Of course, the experimental conditions are largely affected by specimen-dependent factors, such as the kind, shape, and surface roughness. Thus, we need to re-optimize the experimental parameters, in a case where these properties of sample are largely different.
This paper described a method how the focal point of laser irradiation can be automatically controlled just on the on-focus position of a sample in LIBS, based upon a fast variation in the curvature of an electrically-tunable plano-convex lens. The lens can change the focal length with a long working distance and a rapid response time, and the focal length can be periodically varied with a function generator. A Q-switched Nd: YAG laser system is slightly modified by using the tunable lens. A key of this measurement is that the laser beam can be flexibly focused on the specimens even if their shapes are changed every second. A piece of scrapped stainless steel having a tilted surface with some roughness was investigated as a test specimen, where the LIBS plasma could be generated uniformly and firmly along a laser trace when the laser beam was scanned on the specimen surface. The driving frequency of the tunable lens, which controlled a repetition period of the laser irradiation, was an important experimental parameter; in our measurement, the driving frequency was optimized to be 10 Hz when the scan rate of laser was fixed at 3.0 mm/s. The LIBS apparatus suggested in this study slightly modifies a conventional one, and thus it is much less expensive than commercial LIBS apparatuses having complicated optics to control the focal point of laser on the three-dimensional position of a sample. Accordingly, this LIBS system may be applicable to actual on-site/in-line analysis in material production, when specimens flow along a line continuously, like a continuous casting process, or when many specimens such as scrapped pieces are arranged in a line by any mechanical method.
This research was conducted under the support of a Grant-in-Aid from the Ministry of Education, Science, Sports and Culture of Japan. (No. 17 H01903). The authors are grateful to Nippon Steel Corp., Japan, for a Grant for LIBS research (2019).
