Anal. Chem. 2006, 78, 7432-7439
Spatially Resolved, Laser-Induced Breakdown Spectroscopy, Development, and Application for the Analysis of Al and Si in Nickel-Based Alloys Suh-Jen Jane Tsai,* Shi-Yang Chen, Yu-Sheng Chung, and Pai-Chung Tseng†
Department of Applied Chemistry, Providence University, Taichung Hsien, Taiwan, R.O.C., and Department of Mechanical Engineering, National Chung-Hsing University, Taiwan, R.O.C.
Spatially resolved laser-induced breakdown spectroscopy (SRLIBS) was used for the characterization of aluminum and silicon in nickel-based alloys. The very low invasiveness of the technique was one of the figures of merit of LIBS; however, the relative complexity of the instrument often hindered the widely acceptance of LIBS. Spatially resolved LIBS could provide accuracy and precision comparable to those obtained with temporally resolved LIBS (TRLIBS). In the nongated spatially resolved LIBS, the maximum atomic emission could be obtained with relative low continuum background emission at optimum observation spatial position. The study was done with a Nd:YAG laser at 532 nm, 3.0 mJ laser energy, and 0.2 mbar in argon. The experimental results obtained under optimum conditions were compared to those obtained with TRLIBS. SRLIBS gave reliable results without the tedious optimization of the delay time and gate width. Laser-induced breakdown spectroscopy (LIBS), also called laser plasma spectroscopy, has gained much attention in direct analysis,1,2 especially direct solid microanalysis on steel samples;3 bulk analysis of steel samples,4 direct composition measurements of manufactured steel plates,5 and real-time monitoring of hightemperature corrosion in stainless steels,6 to name a few. The temporal and spatial evolution of laser-induced plasma from a steel target was also studied.7 Recently, an impressive work for the quantitative analysis of low-alloy steel was performed by microchip LIBS.8 Generally, LIBS offered acceptable detection limit (micrograms per gram or better in solid). Spatially resolved laser-induced breakdown spectroscopy (SRLIBS) is a relatively new method for metal analysis due to its advantages of being less expensive and less time-consuming. Schechter and co-workers have published † National Chung-Hsing University. (1) Scaffidi, J.; Angel, S. M.; Cremers, D. A. Anal. Chem. 2006, 78, 25-32. (2) Lui, S. L.; Cheung, N. H. Anal. Chem. 2005, 77, 2617-2623. (3) Bleiner, D. Spectrochim. Acta, Part B 2005, 60, 49-64. (4) Sturm, V.; Vrenegor, J.; Noll, R.; Hemmerlin, M. J. Anal. At. Spectrom. 2004, 19, 451-456. (5) Orzi, D. J. O.; Bilmes, G. M. Appl. Spectrosc. 2004, 58, 1475-1480. (6) Garcı´a, P. L.; Vadillo, J. M.; Laserna, J. J. Appl. Spectrosc. 2004, 58, 13471352. (7) Corsi, M.; Cristoforetti, G.; Hidalgo, M.; Iriarte, D.; Legnaioli, S.; Palleschi, V.; Salvetti, A.; Tognoni, E. Appl. Spectrosc. 2003, 57, 715-721. (8) Lopez-Moreno, C.; Amponsah-Manager, K.; Smith, B. W.; Gornushkin, I. B.; Omenetto, N.; Palanco, S.; Laserna, J. J.; Winefordner, J. D. J. Anal. At. Spectrom. 2005, 20, 552-556.
7432 Analytical Chemistry, Vol. 78, No. 21, November 1, 2006
many impressive works on spatially resolved LIBS in the past years. By converting spatial to pseudo temporal resolution in LIBS with simultaneous multifiber spectroscopy, Bulatov et al.9 proved that the spatially resolved LIBS could effectively provide experimental results that are identical to those obtained with optimum delay time. In other words, the temporal gated technique could be replaced by SRLIBS with a less expensive nongated detector, since an expensive pulse-generator was not required. In addition, the time-consuming process for the optimization of the delay time and gate width could be omitted. In SRLIBS, the detection was performed at various distances perpendicularly above the sample surface to avoid the relatively intensive continuum background emission generated from laser-generated plasma. The narrow solid angle observation performed by a nongated detector was similar to the observation performed by a gated detector that could observe the whole plasma; however, an internal standard method was required for obtaining reliable quantitative results due to the matrix effect under the experimental conditions used by Bulatov et al. The properties of the laser-generated plasma were highly dependent on the laser energy irradiance, atmosphere gas, gas pressure, and physical and chemical properties of the sample; the spatial distribution of the emitting atoms also varied with these parameters.10 The spatial distribution of the emission atoms along the axis that was perpendicular to the target varied with the composition of the target and the interaction between the laser and the target. The spatial measurements of the emission atoms and the plasma temperature can also be related to the delay time due to the expansion of the laser-induced plasma. Since the spatial distribution of the emitting atoms varied differently, the detection at different spatial regions of the plasma was similar to the detection with various concentrations of the target element.11 Noll et al. studied the spatial and temporal development of the plasma geometry with a high-speed electrooptic camera. Their work proved that the expansion function z was proportional to a function of time, where zRt(0.76(0.2) and zRt(0.44(0.04) for the times up to 25 ns and after 25 ns, respectively. This implied that the emission spectra of the analyte could be obtained with either an (9) Bulatov, V.; Krasniker, R.; Schechter, I. Anal. Chem. 2000, 72, 2987-2994. (10) Tognoni, E.; Palleschi, V.; Corsi, M.; Cristoforetti, G. Spectrochim. Acta, Part B 2002, 57, 1115-1130. (11) Krasniker, R.; Bulatov, V.; Schechter, I. Spectrochim. Acta, Part B 2001, 56, 609-618. 10.1021/ac060749d CCC: $33.50
© 2006 American Chemical Society Published on Web 09/26/2006
optimized distance from the solid sample surface or with an optimized delay time.12 This observation was consistent with the work of Bulatov et al. 9 LIBS used to be performed under ambient gas of 1 atm.13,14 Under such conditions, plasma with relatively high electron density and temperature was generated. Consequently, quantitative analysis with moderate sensitivity and narrow linear calibration ranges resulted due to an intensive continuum background emission, spectrum-broadening effect and strong self-absorption.15,16 The accuracy and precision of the results obtained by LIBS with a gated detector were greatly dependent on the delay time and the gate width. The internal standard method can be carried out with an optical multichannel analyzer (OMA);17 however, the internal standard method could only provide semiquantitative results for solid analysis without standard reference materials. With a newly designed vacuum ultraviolet Echelle system, microanalysis by laser-induced plasma spectroscopy was accomplished.18 Timeintegrated, spatially resolved LIBS enabled achievement of better analytical results by reducing the interferences from the intensive continuum background emission and the Stark effect.19 Emission spectra obtained with a picosecond laser pulse decayed more rapidly and showed significantly lower background emission than those obtained with a nanosecond laser pulse; however, nanosecond excitation gave a larger radiation intensity due to a larger plasma.20 A relatively new method has been investigated in this work for the determination of aluminum and silicon content in nickelbased alloys by spatially resolved laser-induced breakdown spectroscopy. Actually, the high-temperature alloys or so-called superalloys are important materials for high-tech industrials. They are often used in gas turbine engines and in the hot section for power industrials. Aluminum increases the anticorrosion property of the alloys, and silicon strengthens the hardness and antioxidation property. The challenges in the development of SRLIBS for the analysis of aluminum and silicon were discussed. In addition, the analytical results obtained with spatially resolved LIBS were compared with those obtained with temporally resolved LIBS. EXPERIMENTAL SECTION Experimental Setup. Figure 1 shows the schematic presentation of the SRLIBS setup built up in this work. A Q-switched Nd: YAG laser at 532 nm (Continuum, model Powerlite 9010) was used for generating the sample plasma. The pulse width, repetition rate, (12) Noll, R.; Sattmann, R.; Sturm, V.; Winkelmann, S. J. Anal. At. Spectrom. 2004, 19, 419-428. (13) Russo, R. E.; Mao, X.; Liu, H.; Gonzalez, J.; Mao, S. S. Talanta 2002, 57, 425-451. (14) Bings, N. H.; Bogaerts, A.; Broekaert, A. C. Anal. Chem. 2004, 76, 33133336. (15) Rusak, D. A.; Castle, B. C.; Smith, B. W.; Winefordner, J. D. Crit. Rev. Anal. Chem. 1997, 27, 257-290. (16) Stavropoulos, P.; Palagas, C.; Angelopoulos, G. N.; Papamantellos, D. N.; Couris, S. Spectrochim. Acta, Part B 2004, 59, 1885-1892. (17) Bassiotis, I.; Diamantopoulou, A.; Giannoudakos, A.; Roubani-Kalantzopoulou, F.; Kompitsas, M. Spectrochim. Acta, Part B 2001, 56, 671-683. (18) Radivojevic, I.; Haisch, C.; Niessner, R.; Florek, S.; Becker-Ross, H.; Panne, U. Anal. Chem. 2004, 76, 1648-1656. (19) Bengoechea, J.; Kennedy, E. T. J. Anal. At. Spectrom. 2004, 19, 468-473. (20) Eland, K. L.; Stratis, D. N.; Lai, T.; Berg, M. A.; Goode, S. R.; Michael Angel, S. Appl. Spectrosc. 2001, 55, 279-285.
Figure 1. Schematic representation of SRLIBS setup used in this work. M1 and M2 are the high-power solid-state laser mirrors.
and pulse energy were 4-8 ns, 6 Hz, and 3.0 mJ. The pulse energy output was measured with a pyroelectric joulemeter (Molectron, model PM500, ranges from 2.0 mJ to 20 J with a voltage responsitivity of 9.69 V J-1). The experiments were carried out in argon at 0.2 mbar. The laser beam was guided by two high-power solid-state laser mirrors (M1 and M2) and focused, at normal incidence, on the sample surface with a biconvex quartz lens with a focal length of 100 mm and f number of 4. The focusing lens was mounted on a vertical translation stage so that the lenstarget distance could be changed. The beam diameter was measured by displacing a slit through the focal position. With this method, in the focal point, the laser spot diameter was estimated to be ∼270 µm. The power irradiance was 0.87 GW cm-2. Plasma light collection was performed with a quartz planoconvex lens with focal length of 38.0 mm and f number of 3.8 into the entrance slit of a 0.275-m-focal-length Czerny-Turner spectrograph (Acton Research Corporation, Inc., Spectra Pro-275 with three indexable gratings of 300, 1200, and 2400 grooves mm-1). Light was dispersed using a 2400 grooves mm-1 grating. The reciprocal linear dispersion was 2.5 nm mm -1. The entrance slit was 30 µm wide, and the height was 4.0 mm. Samples. The standard reference materials of high-temperature, nickel-based alloys were analyzed. UNS N06600 (Inconel 600, Series C), UNS N07041 (Rene´ 41, Series C), and UNS N007750 (Inconel X-750, Series C) were products of Claxton Standards, Materials Analysis (Dallas, TX). The standard reference materials of 211X Ni/Cr/Al (11221 D) were purchased from MBH Analytical Ltd. (England). BS 718A (nickel alloy 718) and BS H-3A (Hastelloy X) were purchased from Brammer Standard Company. SRM of 1244 (Inconel 600) and C2402 (high-temperature alloy, Hastelloy C) were purchased from the National Institute of Standard and Technology. Table 1 gives the certified values of aluminum and silicon in the standard reference materials used in this work. Argon of 99.996% purity was used. Data Acquisition. The spatially and temporally resolved LIBS were operated differently. In spatially resolved LIBS, the laser controller system triggered the laser system and ICCD21 (Prince(21) Sabsabi, M.; He´on, R.; St-Onge, L. Spectrochim. Acta, Part B 2005, 60, 12111216.
Analytical Chemistry, Vol. 78, No. 21, November 1, 2006
7433
Table 1. Certified Values of Al, Si, and Nb in the Standard Reference Materials of Nickel-Based Alloys nickel-based alloya
Al (%)
Si (%)
Nb (%)
SRM 1244 SRM C2402 211X Ni/Cr/Al UNS N007750 BS H-3A UNS N06600 UNS N07041c BS 718Ac
0.26 (
