2382
Anal. Chem. 1983, 55, 2382-2387
(24) Riley, J. P.; Chester, R. "Introductionto Marine Chemistry";Academic:
London, 1971.
(25) Grasshoff, K. Deep-sea Res. 1064, 1 1 , 597.
RECEIVED for review June 20,1983. Accepted September 19,
1983. This work was supported by Office of Naval Research Contract N00014-82-K-0740 to K.S.J. and R.L.P., and by a Post-Doctoral Fellowship to J.T. from the Deutsche Forschungsgemeinschaft.
Energy Considerations in Dual Laser Ionization Processes in Flames F. M. Curran,' K. C. Lin? G . E. Leroi, P. M. Hunt, and S. R. Crouch*
Department of Chemistry, Michigan State University, East Lansing, Michigan 48824-1322
Dual laser Ionization (DLI) in flames is a promising technique for trace metal analysis and flame temperature estimation. I n this paper, the success of DLI as a means of enhancing ionization signals arising from resonant excitation of anaiytes Is demonstratedfor LI, Na, Ca, and Sr. I n addition, some DLI schemes which produce no enhancement are exhibited. The lack of enhancement Is explained in terms of two factors: the decline of photoionization cross sections with increasing energetic overshoot into the electronic continuum, and the competition between photoionization and collisional ionization. Considerations applicable to the optimization of DLI experiments are discussed.
Among the many recent laser-based analytical techniques, laser-enhanced ionization has proven particularly useful in the analysis of trace metals in flames (1-21). A closely related technique, dual laser ionization (DLI), is being explored in OUT laboratory. In this technique, two laser beams that overlap both temporally and spatially are employed. The first is a dye laser tuned to resonance with an excited state of the analyte atom; the second beam is used to produce photoionization from the excited state. As reported previously, DLI in flames holds the potential for substantial improvements in analyte detection under some circumstances (22) and also permits the nonphotometric determination of flame temperatures (23). In this paper, new experimental results with a variety of low ionization potential elements are reported. The use of the second (ionizing) laser can yield DLI signals 2 to 3 orders of magnitude larger than the signals obtained with the resonantly tuned dye laser alone. (Unless otherwise indicated, the abbreviation "LEI" is used below for singlewavelength laser-enhanced ionization.) To achieve such enhancements, it is necessary to choose an excitation scheme in which photoionization from the excited state dominates collisional ionization. In this paper, a review of applicable theory is used to elucidate the efficacy-determining aspects of DLI ionization schemes exemplified in experiments with Li, Na, K, Ca, Sr, In, and Cs. Photoionization in rarefied gases through similar pathways has been used by Hurst and co-workers (24), in a method Present address: NASA Lewis Research Center, Cleveland, OH 44135.
*Present address: Department of Chemistry, Cornel1 University, Baker Laboratory, Ithaca, NY 14853. 0003-2700/83/0355-2382$0 1.5010
Table I. Laser Characteristics A. Excitation Beam, Hansch Design, Nitrogen-Pumped Dye Laser; Typical Peak Power at Analytical Line, 5 kW laser dye
approx wavelength range, nm
PBBO a Stilbene 420a Coumarin 460a Rhodamine 6G DCM a
395-410 410-440 450-480 565-605 630-710
B. Ionizing Beam, Nitrogen Laser (Model 0.5-150, NRG, Inc., Madison, WI) Typical Power in Ionizing Beam, 100 kW Wavelength, 337.1 nm (29 665 cm-') a Dyes obtained from Exciton Chemical Co., Dayton, OH. Dye obtained from Eastman Kodak Co., Rochester, NY.
called resonance ionization spectroscopy (RIS). This technique can be extremely sensitive, allowing single atom detection under proper experimental conditions. The DLI technique is basically the extension of RIS to flames, although two separate laser beams are employed rather than the single dye laser beam commonly used to produce both excitation and ionization in RIS (24). Neither LEI nor DLI in flames has yet been applied to routine analysis in typical analytical laboratories, but LEI has found practical application in some difficult trace metal determinations in which extremely high sensitivity is required (e.g., ref 17). The broad spectral range and multiple beams obtainable with new laser technology motivate the development of the DLI method to at least a similar level of practical applicability. If the proper criteria are met, DLI signals can be obtained which are comparable in magnitude to LEI signals obtained a t significantly higher dye laser powers, or longer pulse lengths. The choice of DLI over LEI need not introduce much experimental complexity, since the second (ionizing) laser beam can be derived from the dye laser pumping source. The experimental details of the DLI studies reported herein are reviewed in the next section. In the ionization mechanisms section, various common mechanisms for analyte ionization following one- and two-photon excitation are described, and pertinent rate equations are presented. Following this, results of DLI experiments with seven low ionization potential elements are reported and the widely varying enhancements of 0 1983 American Chemical Society
ANALYTICAL CHEMISTRY, VOL. 55, NO. 14, DECEMBER 1983 2383 Laser-Assisted Ionization Mechanisms A
B
D
C
E
F
section for collisional ionization from level 1, the term (Ei El) is the energy difference between the ionization continuum (IC) and the level 1 (hereafter referred to as the energy defect, If charge recombination is negligible, the total rate of ionization of a specific species in the flame can be expressed as
a),and other symbols take on their usual meanings.
I
”szr
1
Wmaly
Phdon Excnelon,
Stewire Two-photon Excitdion.
COIISI~I
Cdlml
Asslrted slnglc Photon
lonmtb;
lOnlzation
Emtah011.
dni/dt = Ckijcnj J
Two-photon Excltotlm. Collirionol lonmtion
Two-Pep Two-ohdon PhObio11zOWl PhatOlOnlzdDn (Resonance (Off4eSonont) ionhotion.OU1
CollnIOKIl imitation
Figure 1. Common ionization pathways
In laser atomic SpectroscoPY.
DLI signals over LEI signals are traced to their quantum mechanical source. Possible extensions of the reported work are also considered in the final section.
EXPERIMENTAL SECTION A complete description of the DLI experimental system used in this study has been given previously (22,25). In all cases, the resonant laser radiation was produced by a Hiinsch-type dye laser pumped by a nitrogen laser (Model 0.5-150, NRG, Inc., Madison, WI), which also provided the ionizing radiation. The dyes used and typical laser powers are given in Table I. A circular M6k6r burner was used to produce central and mantle hydrogenoxygen-argon flames. The analyte-containing solution was nebulized into the central flame by a crossed flow nebulizer, which has been described elsewhere (25,26).The ion signal was collected with a pair of biased nichrome probes (0.64 mm diameter) which were inserted directly into the flame. The ion signal was amplified by a fast current amplifier (LH0032CG, National Semiconductor, Santa Clara, CA). The circuit for this amplifier was similar in design to that described by Havrilla and Green (27). The amplified output was processed by a boxcar integrator (Model 162/164,EGG-PARC, Princeton, NJ), and the results were displayed on a stripchart recorder. Most solutions were prepared from reagent grade chemicals dissolved in distilled, deionized water. Na, Li, K, and Ca solutions were made from their chloride salts while Cs solutions were made with CsEr and Sr solutions were made from the nitrate salt. The In solution was diluted from a commercial atomic absorption standard (Aldrich Chemical Co., Inc., Milwaukee, WI). Charge sheath formation about the cathode has been observed in both DLI (26) and LEI (28,29)experiments. These sheaths result in reduced ion collection efficiencies at high analyte concentrations. Care was taken in this study to use analyte concentrations which produced negligible charge sheathing effects as actual signal magnitudes were to be compared and thus collection circuit-limited signals were to be avoided.
IONIZATION MECHANISMS Ionization pathways common to laser atomic spectrometry in flames are shown in Figure 1. These can be grouped into two general classes: laser-assisted collisional ionization and direct photoionization. Schemes A-D fall into the first class and have been explored in the laser-enhanced ionization experiments performed to date. Scheme E is of principal interest in DLI. A brief description of collisional processes in flames precedes consideration of the laser ionization schemes. The rate constant for collisional ionization from a given analyte energy level, 1, in a flame has been developed by Lawton and Weinburg (30) and can be expressed as
under the assumption that the reduced mass of the collision pair, p, can represent an average over all collision partners. In eq 1, n, is the number density of the collision partner, T is the flame temperature, QxlC is the partner-averaged cross
where ki; is the rate constant for the production of ions from level j and n, is the number density of level j . In the absence of any optical input, if a Boltzmann distribution of level populations is assumed, it may be seen that there are two competing exponentials involved in the ionization process. Ionization from high-lying levels is exponentially favored as the energy defect is decreased. The number density of these states, however, decreases exponentially with the state energy. If a strong optical field, e.g., that produced by an organic dye laser, is tuned to a resonant transition of the analyte which terminates in state 1, such that for all levels j # 1, Ik$n,l
