Optimization and Calibration of Laser Ablation±Inductively Coupled Plasma Atomic Emission Spectrometry by Measuring Vertical Spatial Intensity Profiles XIANGLEI MAO AND RICHARD E. RUSSO* L awrence Berkeley National L aboratory, Berkeley, CA 94720, USA Vertical spatial emission intensity proÆles for ICP-AES were these studies.29,30 In this paper, the vertical spatial emission measured to optimize and calibrate laser ablation sampling. intensity proÆles of ICP-AES were measured for laser ablation Laser ablation sampling and laser ablation plus liquid and liquid nebulization sampling.The inØuence of carrier gas nebulization sampling were studied. The position of maximum Øow rate and rf power on laser ablation sampling with the ICP emission intensity above the rf load coil changes with gas ICP is discussed in the Ærst part of this paper. Øow rate for both cases, with the maximum position shifting A concern of laser ablation sampling is preferential ablation to higher regions in the plasma at higher Øow rates.The of volatile elements from multicomponent samples. Preferential maximum emission intensity occurs at a Øow rate of ablation depends on the sample properties, laser Øuence, laser approximately 0.2±0.3 l min-1 and at approximately 5±10 mm power density, and laser pulse width.30±37 To determine the above the load coil, which are signiÆcantly different to the extent of preferential ablation for various laser conditions, it values normally employed for liquid nebulization.In addition, is necessary to know the exact composition of the laser ablated by measuring the spatial emission proÆles for laser ablation mass. However, because of preferential ablation, solid stanand nebulization sampling, solutions can be used as standards dards may not provide accurate calibration of the ICP. to calibrate the composition of the laser ablated mass. Solution standards have been proposed for calibration in ICPCalibrated Zn5Cu ratios were measured using UV nanosecond MS23 and ICP-AES.38 For aqueous solution nebulization, the and picosecond laser pulses.Stoichiometric laser ablation analyte dries from liquid droplets to form very small particles sampling of a brass alloy was achieved only by using UV in the plasma. For laser ablation, larger dry particles picosecond laser pulses at high power density. (#1±5 mm) are directly introduced into the ICP. The atomization and excitation processes are expected to be different in Keywords: L aser ablation sampling; inductively coupled plasma these two cases.Because the composition of the ICP and atomic emission spectrometry; vertical spatial proÆle; excitation mechanism will be different for aqueous solution optimization; calibration nebulization and laser ablation sampling, it is prudent to characterize the ICP response for these two cases before using The intensity of an analyte line in ICP-AES is a complex solutions as standards.function of several parameters, including analyte concentration, For ICP-AES, the spectral emission intensity spatial proÆles carrier gas Øow rate, rf power, and the observation height in are governed by ICP electron temperature, electron number the plasma. For nebulized aqueous solutions and electrother- density, total amount of analyte, kinetics of vaporization and mal vaporization, optimum conditions for ICP-AES have been atomization of larger analyte particles, number of incompletely well investigated.1±14 Commonly used conditions are a nebul- evaporated droplets, and excitation mechanism.15±21,39 In a izer gas Øow rate of 0.6±1 l min-1, an rf power of 1.1±1.3 kW, previous study, we demonstrated that the emission spatial and an observation height of 14±18 mm above the load coil.proÆle remained constant for a diverse range of laser power Vertical analyte emission intensity in the ICP is dependent on densities and sample targets.40 However, the vertical spatial the rf power and gas Øow rates.15±21 Solid sample introduction proÆles are not constant for laser ablation and solution nebuliz- using laser ablation provides numerous beneÆts for chemical ation.If solutions are used to calibrate laser ablation sampling, analysis including direct characterization of any solid sample one has to keep all ICP operating conditions constant for and no sample preparation, eliminating dissolution, additional both cases; the total gas Øow, rf power and amount of water solvent waste, and personnel exposure to samples and solvents. must be constant in order to maintain similar excitation Laser ablation requires a smaller amount of sample characteristics and temperature spatial proÆles.Even so, it is (<micrograms) than that required for solution nebulization still possible that the proÆles will be different because of a (milligrams). Because there is no water in the ICP, analyte line different vaporization mechanism for nebulized solution versus intensity may be stronger and molecular background emission laser ablated particles.Solution standards can be used to weaker. The elimination of solvent is also beneÆcial to ICP-MS calibrate laser ablation±ICP-AES only if the emission intensity because of the lower interference of oxide and hydrogen spatial proÆles are the same using both laser ablation and molecular species.22 Analyte excitation behavior during direct liquid nebulization.solid sample introduction will be different to that with nebul- Spatial proÆles for laser ablation and solution nebulization ized aqueous solutions. Previous studies have addressed trans- were measured for different ICP conditions. A parameter range fer tube and ablation chamber designs.23±28 However, to the was established that provided similar emission spatial proÆles best of our knowledge, no study has addressed optimization so that liquid standards could be used to calibrate the ICP.A of the ICP for laser ablation sampling. Such a study is possible calibration graph of Zn5Cu mole ratio versus ICP Zn5Cu only when measuring the entire ICP vertical spatial emission intensity ratio was obtained. The Zn5Cu mole ratio as intensity proÆle. The use of repetitively pulsed laser ablation sampling also provides the necessary precision to carry out a function of laser power density was measured using a Journal of Analytical Atomic Spectrometry, February 1997, Vol. 12 (177±182) 177nanosecond excimer laser (30 ns, 248 nm) and a picosecond The measurement system consists of a monochromator (Spex industries; 270M) with a 1200 grooves mm-1 holo- Nd5YAG laser (35 ps, 266 nm). The inØuence of these laser conditions on accuracy is discussed in the second part of graphic grating and an entrance slit-width of 12.5 mm, and a Peltier-cooled, charged-coupled device (CCD) detector with this paper. 512×512 pixels (EG&G Princeton Applied Research; OMA VISION).The spectral emission from the ICP was imaged EXPERIMENTAL using a quartz lens (5 cm focal length) onto the monochroma- A diagram of the experimental system is shown in Fig. 1. Two tor. This spectrometer simultaneously measures a 30 nm wave- different lasers were used for ablation: a KrF excimer laser length range. The size of the CCD is 1×1 cm. The data from with l=248 nm and a Nd5YAG laser with l=266 nm. The the CCD were digitized and transferred to a microcomputer.pulse durations of the excimer and Nd5YAG lasers are 30 ns For all data reported in this paper, emission intensities are and 35 ps, respectively. Each laser was pulsed at a repetition integrated for 6 s during repetitive ablation. In a previous rate of 10 Hz. The area of the beam was reduced by using paper, the spatial emission proÆle was not inØuenced by the a 6 mm diameter aperture, then focused into the ablation integration time of the CCD detector.40 The brass sample was sample chamber using a plano-convex UV-grade quartz lens pre-ablated for 120 s before ICP intensity measurements to ( f=200 mm). The laser beam spot size at the sample surface achieve enhanced precision.For calibration studies of laser was varied by changing the lens-to-sample distance. The energy ablation with solution nebulization, the experimental pro- of the excimer laser is 25 mJ after the aperture for most of the cedure is to set the ICP power, set the gas Øow rate through experiments.The laser power density at the sample surface is the laser ablation chamber, and introduce water into the ICP 0.07±200 GW cm-2. For the Nd5YAG laser, the energy is from the spray chamber. Without laser ablation, an ICP approximately 2.5 mJ at the sample surface. The power density background emission spectrum proÆle is recorded. The laser is 1±2000 GW cm-2. The laser ablation chamber was mounted is then repetitively pulsed on the sample and the spectral on a xyz micrometre translation stage so that the sampling emission intensity proÆle is measured.The brass solution is spot could be changed after each measurement.29 then introduced into the spray chamber without laser ablation. The samples consisted of small discs (20 mm diameter and The gas Øow rate in the laser ablation chamber was kept 1 mm thickness) of brass. The composition of brass is approxi- constant. With this procedure, the Cu and Zn emission intensity mately 36% Zn and 64% Cu measured by energy dispersive spatial proÆles were obtained with laser ablation and solution X-ray spectrometry.A brass solution was prepared by dissolv- nebulization, respectively, using the same ICP conditions. The ing a portion of this brass sample in nitric acid. Standard procedure is repeated at different gas Øow rates into the laser solutions of Zn and Cu were prepared by dissolving 99.999% ablation chamber. Cu and 99.9% Zn metal, respectively, in dilute HNO3. The Hilbert space distance was used to characterize the The ICP (RFPP; ICP20P) system included a 2.2 kW rf differences between laser ablation sampling and solution nebul- generator, impedance matching network, and mass Øow con- ization emission spatial proÆles.First, the spatial proÆles were trollers (Matheson; 8274). The Øow rate of Ar plasma gas was normalized to their maximum value. Then, f(x) and g(x) were 14 l min-1 and that of the auxiliary gas was 1.0 l min-1. To obtained with maximum values equal to 1 for laser ablation study the inØuence of gas Øow rate on ICP intensity proÆles and solution nebulization, respectively. For functions f(x) and during laser ablation sampling, the carrier gas transports the g(x), the Hilbert space distance (L2 distance) was deÆned as ablated mass directly into the ICP, without a spray chamber.For calibration studies, the exit port of the ablation chamber D= SPb a [f(x)-g(x)]2 dx (1) was connected to a steel T-connector (Swagelok), so that the carrier gas could be mixed with gas from a spray chamber.The Øow rate of gas into the spray chamber was Æxed at where x is the vertical distance and a and b are the boundaries 0.4 l min-1, the lowest level at which stability could be main- for the low and high observation region in the ICP. The tained for this nebulizer. The Øow rate of carrier gas into the Hilbert distance between spatial functions of laser ablation laser ablation chamber was varied from 0.2 to 1.0 l min-1.The and solution nebulization describes the error when solutions mixed gas was introduced into the central channel of the ICP are used as standards for calibration and the spatial proÆles torch. When using laser ablation to transport the sample into do not overlap. the ICP, water was introduced into the spray chamber to maintain constant ICP conditions. RESULTS AND DISCUSSION Optimization of Laser Ablation±ICP-AES The main parameters that inØuence spectral emission intensity in the ICP are carrier gas Øow rate, rf power, and observation height above the load coil (HALC).To optimize these conditions for laser ablation sampling and laser ablation sampling with solution nebulization, the ICP spatial emission intensity proÆles were measured for different gas Øow rates and rf powers. Both sample introduction conÆgurations were used for this study. Fig. 2 shows the Cu 224.7 nm (brass sample) spatial emission intensity proÆles for different carrier gas Øow rates with an rf power of 1500 W using the conÆguration in which the carrier gas Øows through the laser ablation chamber and transports the ablated sample directly into the ICP torch.The Øow rates were 0.1, 0.15, 0.2, 0.3, 0.5 and 1.0 l min-1. Fig. 3 shows the Cu 224.7 nm spatial emission intensity proÆles for different carrier gas Øow rates through the laser ablation chamber with a 0.4 l min-1 Ar Øow rate through the nebulizer while aspirating water.The Øow rates through the laser Fig. 1 Diagram of experimental system for laser ablation±ICP-AES. ablation chamber were 0.2, 0.3, 0.4, 0.6, 0.8 and 1.0 l min-1. 178 Journal of Analytical Atomic Spectrometry, February 1997, Vol. 12improved by a factor of about 3 with a lower Øow rate. For the laser ablation plus water case (Figs. 3 and 5), the intensity is improved by almost a factor of 10. These studies also show that the best HALC for maximum emission is lower than that generally used for conventional liquid nebulization sampling.It is interesting that the maximum intensity is similar for both cases, considering that the ICP excitation conditions are different with and without water. For ICP=1500 W, a maximum intensity of #7.8×105 occurs with a Øow rate of 0.2 l min-1 in the laser ablation chamber plus 0.4 l min-1 through the nebulizer (0.6 l min-1 total), compared with #8.5×105 with a Øow rate of 0.3 l min-1 in the laser ablation chamber.The emission intensity depends on two primary factors: the amount of sample and the excitation conditions of the ICP. For this Æxed laser power density study, the amount of ablated sample was approximately constant for this homogeneous Fig. 2 Vertical spatial emission intensity proÆles of Cu in the ICP at sample; the RSD is approximately 3%. When the carrier gas different gas Øow rates (l min-1). Carrier gas through laser ablation Øow rate increases, transport efficiency improves and the chamber and directly into the ICP.Rf power, 1500 W. amount of mass delivered to the ICP increases.41 However, increasing the Øow rate decreases the excitation temperature of the ICP.17 There are more Ar atoms per unit time in the ICP for higher gas Øow rates. With increasing Ar gas Øow rate, the maximum emission intensity in the ICP decreases and the peak position shifts upward.15±17,19±21 There is a trade off in transport efficiency versus temperature that governs the intensity and spatial properties of these emission data.A complete analysis would require measurement of the ICP temperature behavior versus Øow rate. Monitoring peak intensity is important for studying changes related to Øow rate. However, peak intensity alone does not deÆne analytical sensitivity. The ratio of analyte spectral intensity to continuum emission background is important for optimization of sensitivity.11 To determine the optimum height for AES observation, it is necessary to know the spatial proÆles of both line intensity and background. The ICP background emission decreased with increasing Øow rates for both laser Fig. 3 Vertical spatial emission intensity proÆles of Cu in the ICP at ablation and laser ablation plus water. Figs. 4 and 5 show the different gas Øow rates (l min-1). Carrier gas through laser ablation spatial proÆles for the Cu intensity ratio to background for chamber with a 0.4 l min-1 gas Øow through spray chamber with laser ablation and laser ablation plus water, respectively. water nebulization.Rf power, 1500 W. Comparing Fig. 4 with Fig. 2, the maximum position (HALC) for the ratio is approximately 5 mm higher than that of absolute intensity. The absolute intensity at 0.2 l min-1 is The intensities shown in Figs. 2 and 3 were backgroundcorrected. The laser power density used in these experiments approximately three times stronger than that at 1.0 l min-1; the ratio is almost the same.The maximum ratio occurs at a was 0.9 GW cm-2 and the brass samples were pre-ablated for each measurement. Therefore, the ablated mass was approxi- Øow rate of 0.5 l min-1, whereas maximum intensity occurs at 0.2 l min-1. For laser ablation plus water, the absolute intensity mately constant for each of these proÆles. The Cu intensity was measured at Æve separate spots on the brass sample for at 0.2 l min-1 is approximately ten times stronger than that the 0.3 l min-1 Øow rate to determine the error associated with these measurements, primarily due to the laser ablation sampling.The error at this Øow rate is approximately 3% RSD for the Cu measurements and is consistent with previous studies using repetitive pulsing and pre-ablation.29 As expected, the position of maximum ICP emission intensity changes with gas Øow rate for both conÆgurations, with the maximum shifting to higher regions in the plasma at higher Øow rates.The Zn emission intensity spatial proÆles behaved similarly to those of Cu and are therefore not shown. The Cu peak intensity initially increases with gas Øow rate, reaches a maximum, and then decreases in the laser ablation conÆguration (Fig. 2). The maximum occurs at a Øow rate of approximately 0.2±0.3 l min-1. For laser ablation plus water, the Cu intensity decreases with increasing gas Øow rate above 0.2 l min-1 in the laser ablation chamber (with a constant 0.4 l min-1 gas Øow through the nebulizer) (Fig. 3).Increasing the ICP power increases the ICP intensity, but the maximum analyte intensity always occurs at 0.2±0.3 l min-1 for both cases. This Øow rate Fig. 4 Ratio of Cu line intensity to ICP background as a function of is considerably lower than that generally accepted for optimum height above load coil at different gas Øow rates (l min-1). Carrier performance using liquid nebulization alone. For the laser gas through laser ablation chamber and directly into the ICP.Rf power, 1500 W. ablation case (cf. Figs. 2 and 4), the emission intensity is Journal of Analytical Atomic Spectrometry, February 1997, Vol. 12 179of vertical height (temperature) in the ICP. For laser ablation, the analyte already exists in the solid particle or atomic form; melt ejection or sublimation occurs from the solid and there is no aqueous evaporation. Measurement of emission spatial proÆles provides a good way to determine the conditions under which standard solutions can be used for calibration.From eqn. (2), the intensity proÆle I(x) is a function of the temperature proÆle T (x), and the number density of atoms or ions g(x). If T (x) or g(x) in laser ablation and solution sampling is different, I(x) will be different. Therefore, the ICP conditions must be established to ensure that I(x) is the same for both laser ablation and solution nebulization. The carrier gas Øow rate and rf power can be optimized to minimize the difference between laser ablation and solution nebulization. For this study, the spatial proÆles for laser ablation plus water versus brass (Cu+Zn) solutions were compared.Fig. 6(a) and (b) shows the normalized Cu emission intensity spatial proÆles for Fig. 5 Ratio of Cu line intensity to ICP background as a function of height above load coil at different gas Øow rates (l min-1). Carrier laser ablation plus water at 1.0 and 0.2 l min-1 gas Øow rates, gas through laser ablation chamber with a 0.4 l min-1 gas Øow through respectively.The nebulizer Øow rate was 0.4 l min-1 in both spray chamber with water nebulization. Rf power, 1500 W. cases. The total Øow was always the Øow through the ablation chamber plus the nebulizer; 1.4 l min-1 in Fig. 6(a) and 0.6 l min-1 in Fig. 6(b). When the ablation chamber Øow rate at 1.0 l min-1; the line-to-background ratio is approximately is 1.0 l min-1, the spatial proÆles are different [Fig. 6(a)], even four times better. The maximum ratio and absolute intensity though the total water loading into the ICP is the same. The occur at the same Øow rate (0.2 l min-1 through the laser different `particle' behavior in the ICP is apparent under these ablation chamber and 0.4 l min-1 through the spray chamber). conditions. The spatial proÆles using a 0.2 l min-1 Øow rate Analyte and background intensity are increased at lower Øow in the ablation chamber are almost identical for both laser rates, and optimum sensitivity is a balance between the two.ablation plus water versus nebulized standard solutions If absolute intensity is used for optimization, a carrier gas [Fig. 6(b)]. Under these Øow conditions, the difference in Øow rate of 0.2±0.3 l min-1 is best for both laser ablation and `particle' behavior is minimized. laser ablation plus water sampling. The observation height The Hilbert space distance was used to quantify the difference should be 1±5 mm and 3±8 mm above the load coil for laser between two emission intensity proÆles; the smaller the value, ablation and laser ablation plus water, respectively.As regards the more identical the proÆles. The difference between laser both net intensity and line-to-background ratio, the best Øow rate is 0.2±0.3 l min-1 and the best observation height is 1±5 mm for laser ablation sampling. For laser ablation plus water, 0.2 l min-1 provides the best performance with our nebulizer at 0.4 l min-1.The best observation height is 5±11 mm above the load coil. All of these observations were based on Cu and Zn emission lines located between 200 and 230 nm. Other element lines are expected to have different proÆles. However, the research indicates that lower gas Øow rates and lower observation heights provide enhanced sensitivity for laser ablation sampling. Calibration of Laser Ablation Sampling with Solution Standards To calibrate laser ablation sampling, the composition of the sample introduced into the ICP has to be known.The ICP line intensity can be expressed as:42 I(x)=Ahc 4pBAigi liQ g(x) e-Ei/[kBT(x)] (2) where Ai is the Einstein transition probability for spontaneous emission, Ei the energy of the transition, gi the statistical weight of level Ei, T the excitation temperature, li the wavelength, g the emitting atom density, h Planck's constant, kB Boltzman's constant, c the speed of light and Q the internal partition function.In order to calibrate laser ablation using solution standards, the temperature proÆles have to be the same for both laser ablation sampling and solution nebulization. Therefore, for laser ablation the same amount of water and acid must be nebulized into the ICP. However, even with the same amount of Ar gas and water, it is still possible that excitation conditions and vertical spatial emission proÆles may Fig. 6 Normalized Cu emission intensity in the ICP as a function of not be similar because of different particle sizes and different height above load coil with gas Øow rates of (a) 1.0 and (b) 0.2 l m-1, vaporization behavior.For liquid nebulization, aqueous drop- in the laser ablation chamber and 0.4 l min-1 through the spray lets undergo evaporation, atomization, ionization, excitation chamber. Solid line is laser ablation sampling plus water. The broken line is for a brass solution through the nebulizer without laser ablation.and recombination processes. These processes are a function 180 Journal of Analytical Atomic Spectrometry, February 1997, Vol. 12ablation and solution nebulization of standards was calculated discussion on optimization, emission line intensity was maximum at a gas Øow rate of 0.2 l min-1. A gas Øow rate of for different gas Øow rates and rf powers. The offset in Cu and Zn spatial proÆles with carrier gas Øow rate was measured at 0.2 l min-1 also provided the optimum overlap in spatial proÆles for accurate calibration.From Fig. 7(a) and (b), it can three different ICP powers [Fig. 7(a) and (b)]. The difference (error) for Cu ranges from approximately 0.05 at 0.2 l min-1 to be seen that laser ablation and solution nebulization provided identical spatial emission intensity proÆles using a Øow rate of 0.4 at 1 l min-1. For Zn, the difference is small (#0.02) until the gas Øow rate reaches 0.8 lmin-1, and is not very sensitive 0.2 l min-1.Therefore, these conditions were used to calibrate the Zn5Cu ratio using standard solutions. An analytical to ICP power at low Øow rates. Zn intensity proÆles were not as sensitive as those for Cu when the gas Øow rate changed. A working curve was generated and used to determine the accurate ratio for the brass sample (0.56 mole ratio). possible explanation is that Zn is easier to atomize because of its lower melting- and boiling-point than Cu. Zn may be The calibrated Zn5Cu mole ratio as a function of laser power density is shown in Fig. 8 for ablation with the 30 ns completely atomized for both laser ablation and solution nebulization sampling for gas Øow rates less than 0.8 l min-1. A UV excimer laser. Zn is severely preferentially ablated at lower laser power density, below 0.3 GW cm-2. When the laser lower gas Øow rate and increased temperature may be required to atomize Cu efficiently. In general, the offset in spatial proÆles power density is higher than 0.3 GW cm-2, the ratio is close to accurate.The insert in Fig. 8 shows that even at high power will depend on the particular species, the excitation state and ICP characteristics. Better overlap was measured for both density, the composition of the ablated mass is different to the composition of the solid sample. Similar data have been species at the lower Øow rate. If two intensity proÆles are the same, the Hilbert space measured previously, but not against a calibrated ICP.43 At lower power density the laser ablation process is mainly distance is equal to zero.Because of instrumental errors and Øuctuations in laser ablation and solution nebulization, the thermal; Zn is easier to vaporize than Cu. As the power density is increased, the amount of Cu increases faster than that of Hilbert space distance cannot be zero in these experiments. To determine the instrumental and laser ablation errors, three Zn, because of a higher surface temperature.32,33,43 The stoichiometric ratio is approached as the power density reaches laser ablation and solution ICP proÆles were obtained at a gas Øow rate of 0.6 l min-1 and an rf power of 1500 W.The #0.3 GW cm-2, the roll-off region for mass ablation efficiency previously observed.32,33 There are several competing mechan- error between laser ablation spatial emission proÆles themselves is from 0.04 to 0.09. The error between repetitive solution isms inØuencing mass removal.When the laser power density reaches 0.3 GW cm-2, plasma shielding may be import- nebulization spatial emission proÆles is about 0.02. The larger values for laser ablation demonstrate the noise from this ant.32,33,43 The plasma is dense and hot enough to absorb a portion of the laser energy. The hot, high pressure plasma will sampling technology. This error represents a `jitter' in the overall ICP behavior to the particle size distribution introduced interact with the solid sample causing sputtering.There is a trade off between direct laser heating and plasma heating that during laser ablation. At a gas Øow rate of 0.2 l min-1, the difference between laser ablation and solution nebulization is will provide the most accurate analysis. A higher laser power density provides better accuracy, although accurate analysis is approximately 0.05. Therefore, for these conditions the spatial proÆles are essentially identical for laser ablation and solution not achieved for the brass sample using the UV nanosecondpulsed laser.The error is approximately 10% at 0.3 GW cm-2. nebulization, since the Hilbert space distance is approximately equal to that of laser ablation sampling alone. From the earlier Fig. 9 shows the calibrated Zn5Cu ratio as a function of laser power density using the picosecond-pulsed laser with l= 266 nm. The behavior is different to that using the nanosecondpulsed laser; the ratio increases with increasing power density and becomes stoichiometric after approximately 10 GW cm-2.The region of accurate analysis, after #10 GW cm-2 is also the region after which roll-off occurs for the picosecond-pulsed laser material interaction.32,33 These data demonstrate that the mechanisms for nanosecond- and picosecond-pulsed laser Fig. 8 Zn5Cu mole ratio from laser ablated mass as a function of Fig. 7 Hilbert space distance between laser ablation and solution laser power density.Solid line is the calibrated ratio from a brass solution. Excimer laser with l=248 nm and pulse duration=30 ns. ICP spatial intensity proÆles versus gas Øow rate in laser ablation chamber. (a) Cu and (b) Zn. Insert shows an expanded scale for the higher power density. Journal of Analytical Atomic Spectrometry, February 1997, Vol. 12 181REFERENCES 1 Sadler, D. A., Littlejohn, D., and Perkins, C. V., J. Anal. At. Spectrom., 1996, 11, 207. 2 Ebdon, L., Foulkes, M., and O'Hanlon, K., Anal.Chim. Acta, 1995, 311, 123. 3 Galley, P. J., Horner, J. A., and Hieftje, G. M., Spectrochim. Acta, Part B, 1995, 50, 87. 4 Kantor, T., and Zaray, G., Microchem. J., 1995, 51, 266. 5 Stroh, A., and Vollkopf, U., J. Anal. At. Spectrom., 1993, 8, 35. 6 Golightly, D. W., and Leary, J. J., Spectrochim. Acta Rev., 1991, 14, 111. 7 Nickel, H., Zadgorska, Z., and Wolff, G., Spectrochim. Acta, Part B, 1993, 48, 25. 8 Boumans, P. W. J. M., in Inductively Coupled Plasma Emission Spectroscopy, Part I: Methodology, Instrumentation, and Perfor- Fig. 9 Zn5Cu mole ratio from laser ablated mass as a function of mance, ed. Boumans, P. W. J. M., Wiley, New York, 1987, ch. 4. laser power density. Solid line is the calibrated ratio from a brass 9 Thompson, M., and Barnes, R. M., in Inductively Coupled Plasma solution. Nd5YAG laser with l=266 nm and pulse duration=35 ps. in Analytical Atomic Spectroscopy, ed. Montaser, A., and Golightly, D. W., VCH, New York, 2nd edn., 1992, ch. 5. 10 Dickinson, G. W., and Fassel, V. A., Anal. Chem., 1969, 41, 1021. 11 Boumans, P. W. J. M., and de Boer, F. J., Spectrochim. Acta, Part B, 1972, 27, 391. ablation are different. For brass analysis, the picosecond-pulsed 12 Scott, R. H., Fassel, V. A., Kniseley, R. N., and Nixon, D. E., laser with l=266 nm provides a wide power density region in Anal. Chem., 1974, 46, 75. which the ablated material has the same composition as the 13 GreenÆeld, S., Jones, I. L., McGeachin, H.M., and Smith, P. B., solid. In previous work, we demonstrated that the picosecond- Anal. Chim. Acta, 1975, 74, 225. 14 Fassel, V. A., and Kniseley, R. N., Anal. Chem., 1974, 46, 1110A. pulsed laser also provides better sensitivity than the nano- 15 Blades, M. W., and Horlick, G., Spectrochim. Acta, Part B, 1981, second-pulsed laser.32 36, 861. 16 Kawaguchi, H., Ito, T., Ota, K., and Mizuike, A., Spectrochim. Acta, Part B, 1980, 35, 199. 17 Kawaguchi, H., Ito, T., and Mizuike, A., Spectrochim.Acta, Part CONCLUSION B, 1981, 36, 615. 18 Koirtyohann, S. R., Jones, J. S., Jester, C. P., and Yates D. A., By measuring vertical spatial emission intensity proÆles, ICPSpectrochim. Acta, Part B, 1981, 36, 49. AES operating conditions can be established for optimum 19 Savage, R. N., and Hieftje, G. M., Anal. Chem., 1980, 52, 1267. response during laser ablation sampling. Compared with the 20 Fister, J. C., III, and Olesik, J.W., Spectrochim. Acta, Part B, ICP operating conditions for nebulized aqueous solutions, 1991, 46, 869. 21 Olesik, J. W., and Den, S. J., Spectrochim. Acta, Part B, 1990, laser ablation sampling provides its best sensitivity at lower 45, 731. carrier gas Øow rates and lower observation heights in the 22 Horlick, G., and Shao, Y. B., in Inductively Coupled Plasma in plasma. The inØuence of rf power on emission intensity was Analytical Atomic Spectroscopy, ed. Montaser, A., and Golightly, not as sensitive as gas Øow rate for the elements investigated.D. W., VCH, New York, 2nd edn., 1992, ch. 12. 23 Cromwell, E. F., and Arrowsmith, P., Anal. Chem., 1995, 67, 131. By measuring vertical spatial emission intensity proÆles, 24 Arrowsmith, P., and Hughes, S. K., Appl. Spectrosc., 1988, 42, 1231. laser ablation sampling with the ICP can be calibrated by 25 Leis, F., and Laqua, K., Spectrochim. Acta, Part B, 1978, 33, 727. using standard solutions. The gas Øow rate should be less than 26 Ishizuka, T., and Uwamino, Y., Anal.Chem., 1980, 52, 125. 0.3 l min-1 when a standard nebulizer system is used with a 27 Carr, J. W., and Horlick, G., Spectrochim. Acta, Part B, 1982, 37, 1. 28 Liu, X. R., and Horlick, G., Spectrochim. Acta, Part B, 1995, 0.4 l min-1 Ar Øow. With calibrated ICP, preferential vaporiz- 50, 537. ation is shown to exist throughout a wide power density range 29 Chan, W. T., and Russo, R. E., Spectrochim. Acta, Part B, 1991, using a UV-pulsed laser.The Zn5Cu ratio from ablation of 46, 1471. brass samples with a UV nanosecond-pulsed laser decreased 30 Arrowsmith, P., Anal. Chem., 1987, 59, 1437. 31 Cromwell, E. F., and Arrowsmith, P., Appl. Spectrosc., 1995, with increasing laser power density, and stabilized for power 49, 1652. densities higher than 0.3 GW cm-2. However, the composition 32 Russo, R. E., Appl. Spectrosc., 1995, 49, 14A. of the ablated mass was always Zn-rich. Using UV picosecond- 33 Shannon, M.A., Mao, X. L., Fernandez, A., Chan, W. T., and laser pulses, the Zn5Cu ratio increased with increasing laser Russo, R. E., Anal. Chem., 1995, 67, 4522. 34 Russo, R. E., Mao, X. L., Chan, W. T., Bryant, M. F., and Kinard, power density and stabilized at an accurate level; the composi- W. F., J. Anal. At. Spectrom., 1995, 10, 295. tion of the laser ablated mass was the same as that of the solid 35 Chan, W. T., Mao, X. L., and Russo, R. E., Appl. Spectrosc., 1992, sample when the laser power density was higher than 46, 1025. 10 GW cm-2. The picosecond-pulsed laser provided better 36 Omori, N., and Inoue, M., Appl. Surf. Sci., 1992, 54, 232. 37 Mochizuki, T., Sakashita, A., Tsuji, T., Iwata, H., Ishibashi, Y., ablation efficiency and better accuracy for chemical analysis and Gunji, N., Anal. Sci., 1991, 7, 479. than the nanosecond-pulsed laser. 38 Baldwin, D. P., Zamzow, D. S., and D'Silva, A. P., Anal. Chem., Standard solutions represent a potential methodology for 1994, 66, 1911. calibrating the ICP for laser ablation sampling. However, it 39 Mostaghimi, J., Proulx, P., Boulos, M. I., and Barnes, R. M., Spectrochim. Acta, Part B, 1985, 40, 153. would be desirable not to use solutions. Additional work is 40 Caetano, M., Mao, X. L., and Russo, R. E., Spectrochim. Acta, necessary to understand laser ablation better so that solids Part B, 1996, 51, 1473. can be used as standards without matrix effects. It may be 41 Rosner, D. E., Mackowski, D. W., Tassopoulos, M., Castillo, J., possible that operating conditions can be established for the and Garcia-Ybarra, P., Ind. Eng. Chem. Res., 1992, 31, 760. 42 Mermet, J. M., in Inductively Coupled Plasma Emission ICP and UV picosecond-laser pulses to provide stoichiometric Spectroscopy, Part II: Application and Fundamentals, ed. Boumans, analysis, in which case solid standards can be trusted for P. W. J. M., Wiley, New York, 1987, ch. 10. calibration. 43 Mao, X. L., Chan, W. T., Caetano, M., Shannon, M. A., and Russo, R. E., Appl. Surf. Sci., 1996, 96±98, 126. This research was supported by the US Department of Energy, Office of Basic Energy Sciences, Chemical Sciences Division, Paper 6/06059E Processes and Techniques Branch, under Contract No. Received September 3, 1996 Accepted October 8, 1996 DE-AC03±76SF00098. 182 Journal of Analytical Atomic Spectrometry, February 1997, Vol. 12