Development of Tunable Picosecond Dye Laser for Multiphoton

A distributed-feedback dye laser has been developed for achieving the efficient multiphoton ionization of chlo- robenzene and dichlorobenzene, that is...
0 downloads 0 Views 78KB Size
Anal. Chem. 2001, 73, 4417-4421

Development of Tunable Picosecond Dye Laser for Multiphoton Ionization of Dioxin Precursors in Supersonic Jet/Time-of-Flight Mass Spectrometry Noriyoshi Yoshida, Yasuyuki Hirakawa, and Totaro Imasaka*

Department of Chemical System and Engineering, Graduate School of Engineering, Kyushu University, 6-10-1 Hakozaki, Higashi-Ku, Fukuoka 812-8581, Japan

A distributed-feedback dye laser has been developed for achieving the efficient multiphoton ionization of chlorobenzene and dichlorobenzene, that is, precursor molecules of dioxins. This tunable picosecond laser with a narrow spectral line width, that is, a nearly transformlimited pulse, provides a more efficient ionization than the nanosecond laser, which is currently in use in supersonic jet spectrometry. The advantage of picosecond over nanosecond and femtosecond lasers is also discussed on the basis of the theoretical model reported in a previous paper. Dioxin is emitted from many incinerators into the atmosphere and accumulates in a wide variety of organisms, including fish and animals and, ultimately, in humans. Dioxin is known to be an extremely toxic compound and is suspected of being mutagenic. To control the combustion process in an incinerator from the viewpoint of reducing the concentration of emitted dioxin, it would be desirable to develop an analytical instrument that allows the on-line real-time monitoring of dioxin and its precursors. At the present time, dioxin is measured by gas chromatography combined with mass spectrometry (GC/MS) after lengthy pretreatment procedures such as extraction, column separation, and concentration. As a result, the procedure is costly (e.g., $3000 for a sample) and lengthy (at least one week, typically a few months), and thus, it is not of practical use in on-line real-time monitoring. Recently, the use of supersonic jet spectrometry combined with multiphoton ionization mass spectrometry (SSJ/MPI-MS) has been proposed to solve this problem.1-8 In fact, this technique has already been applied to on-line real-time monitoring of chlorobenzene emitted from an incinerator.9 This approach allows the measurement of a mass spectrum for an analyte whose * To whom correspondence should be addressed. Phone: +81-92-642-3563. Fax: +81-92-632-5209. E-mail: [email protected]. (1) Syage, J. A. Anal. Chem. 1990, 62, 505A. (2) Weickhardt, C.; Zimmermann, R.; Bosel. U.; Schlag, E. W. Rapid Commun. Mass Spectrom. 1993, 7, 183. (3) Zimmermann R.; Bosel, U.; Weickhardt, C.; Lenoir, D.; Schramm, K. W.; Kettrup, A.; Schlag, E. W. Chemosphere 1994, 29, 1877. (4) Gittins, C. M.; Castaldi, M. J.; Senkan, S. M.; Rohlfing, E. A. Anal. Chem. 1997, 69, 286. (5) Nir, E.; Hunziker, H. E.; de Vries, M. S. Anal. Chem. 1999, 71, 1674. (6) Tzeng, W. B.; Narayan, K.; Lin, J. L. Appl. Spectrosc. 1999, 53, 731. (7) Zimmermann, R.; Bosel, U.; Heger, H. J.; Rohwer, E. K.; Schlag, E. W.; Kettrup, A. J. High Resolut. Chromatogr. 1997, 20, 1. (8) Zimmermann, R.; Rohwer, E. K.; Heger, H. J. Anal. Chem. 1999, 71, 4148. 10.1021/ac010187s CCC: $20.00 Published on Web 08/11/2001

© 2001 American Chemical Society

resonance absorption line coincides with the laser wavelength within an accuracy of 0.01 nm. Therefore, interference by compounds having similar chemical structures and similar spectral properties, but with different toxicities, can be discriminated. However, the concentration of dioxin in incinerator waste gas is very low, and as a result, an analytical instrument with high sensitivity is needed. It has recently been reported that ionization efficiency can be significantly improved by the use of a laser with a short pulse width.10 This can be attributed to the fact that the lifetime of the excited state of an aromatic hydrocarbon that contains numerous chlorine atoms becomes short as a result of spin-orbit interactions. Therefore, it would be predicted that excitation and subsequent ionization could be achieved instantly using a short laser pulse. Generally, a short optical pulse has a wide spectral line width, as is evident from the uncertainty principle. Such a laser, that is, one with a broad line width, provides rather poor ionization efficiency and poor spectral selectivity for a molecule that represents a spectrum consisting of sharp spectral lines. The use of a laser whose pulse width is identical to the lifetime of the analyte molecule was recently suggested.11 Unfortunately, the lifetime of the dioxin molecule is unknown, but it is likely to be in the order of picoseconds. Thus, efficient ionization may be established using a picosecond laser with a narrow line width, that is, a transform-limited pulse in the picosecond region; however, such a laser has not yet been commercially developed. A dye laser consisting of a short cavity provides several sharp spikes in a time domain. It is possible to isolate the first spike using an additional cavity for quenching; however, this provides a laser with a broad spectral bandwidth. To obtain monochromatic light, it is necessary to use a dispersive element, such as a grating, in the cavity; however, it increases the cavity length for an increase of dispersion, thus preventing the generation of a short laser pulse. It is, of course, possible to place a dispersive element outside the cavity; however, the light intensity is substantially reduced, and a complicated amplifier/saturable absorber system is required. An alternative approach for overcoming this problem is the use of a distributed-feedback dye laser.12-18 In this case, two beams, (9) Zimmermann, R.; Heger, H. J.; Blumenstock, M.; Dorfner, R.; Schramm, K.-W.; Boesl, U.; Kettrup, A. Rapid Commun. Mass Spectrom. 1999, 13, 307. (10) Matsumoto, J.; Lin, C. H.; Imasaka, T. Anal. Chem. 1997, 69, 4524-4529. (11) Matsumoto, J.; Imasaka, T. Anal. Chem. 1999, 71, 3763. (12) Bor, Zs. IEEE J. Quantum Electron. 1980, 16, 517.

Analytical Chemistry, Vol. 73, No. 18, September 15, 2001 4417

separated from a single beam by a grating, are superimposed on the surface of the dye solution to produce an interference pattern. The emitting wavelength, λL, can be calculated by the following equations.

λL ) 2nΛ

(1)

Λ ) λL/2 sin θ

(2)

where n is the refractive index of the medium; Λ, the fringe spacing; and θ, the angle from the normal surface. A unique feature of the distributed-feedback dye laser is its short cavity length (essentially no cavity), which results in a short laser pulse. The laser wavelength is determined by the fringe spacing, which, in turn, is determined by the angle of incidence of the pump beam. Hence, no dispersive element, such as a grating, is necessary inside the cavity. However, the emitting wavelength is strongly affected by the change in the temperature of the dye solution, because the value of the refractive index, which appears in eq 1, is substantially changed during the operation of the laser. This represents the probable reason that this type of laser has not yet been used in spectroscopic measurements and has not yet been commercialized. In this study, we describe the construction of a distributedfeedback dye laser for the generation of a nearly transform-limited pulse in the picosecond region, which is combined with a wavemeter for continuous monitoring of the emitting wavelength. This laser system can be employed for the excitation and subsequent ionization of chlorobenzenes in supersonic jet/timeof-flight mass spectrometry. We also demonstrate herein a more efficient ionization using a picosecond laser, rather than a nanosecond one. We also report on the potential advantages of the picosecond laser having a narrow spectral line width, that is, a Fourier transform-limited pulse in the picosecond region, for the efficient multiphoton ionization of dioxin molecules based on the theoretical model reported in a previous paper.11 EXPERIMENTAL SECTION Distributed Feedback Dye Laser. Figure 1 illustrates a schematic diagram of the experimental apparatus. The third harmonic emission of a Nd:YAG laser (Coherent, Infinity, 355 nm, 10 Hz, 3 ns) is used as the pump source for a distributed-feedback dye laser, which is composed of an oscillator and two amplifiers. In the oscillator stage, the pump beam is split into two parts using a grating (2400 lines/mm), and these are then combined by reflection using two aluminum mirrors to form an interference pattern on the surface of the dye solution. The laser dye used in this study was Coumarin 153 (Exciton) at a concentration of ∼3.4 × 10-3 M. The temporal profile of the laser pulse was measured using a streak camera (Hamamatsu, C4334, ∆t ) 15 ps), and the emitting wavelength, and the spectral line width, by a wavemeter (Burleigh, WA5500; accuracy, 0.002 nm; resolution, 0.0001 nm). (13) Bor, Zs.; Muller, A.; Racz, B.; Scha¨fer, F. P. Appl. Phys. B 1982, 27, 77. (14) Szabo, G.; Bor, Zs. Appl. Phys. B 1983, 31, 1. (15) Jasny, J. Optics Comm. 1985, 53, 238. (16) Suesse, K. E.; Weidner, F. Appl. Phys. B 1990, 51, 267. (17) Mu ¨ ller, A. Appl. Phys. B 1996, 63, 443. (18) Zitelli, M.; Fazio, E.; Bertolotti, M. IEEE J. Quantum Electron. 1998, 34, 609.

4418 Analytical Chemistry, Vol. 73, No. 18, September 15, 2001

Figure 1. Experimental apparatus.

The beam from the oscillator passes through two dye cells for subsequent amplifications. The beam from the final amplifier is focused on a potassium dihydrogen phosphate (KDP) crystal to generate the second harmonic emission. The resulting ultraviolet beam is then used for multiphoton ionization of chlorobenzenes in a supersonic jet. Supersonic Jet Spectrometer. A sample placed in a reservoir is heated and evaporated for injection from a pulsed nozzle into a vacuum, using argon as the carrier gas, and is ionized by the distributed-feedback dye laser. The ions are accelerated by a high potential toward the same direction to jet expansion and are detected by an assembly of microchannel plates. This collinear ionization scheme is used to enhance the sensitivity. The output signal, that is, the mass spectrum, is recorded by a digital oscilloscope. The resolution of the mass spectrometer is 210 at m/z ) 94. Details of the experimental apparatus are described elsewhere.19,20 Monochlorobenzene and m-dichlorobenzene were used as samples without further purification. RESULTS AND DISCUSSION Dye Laser. Figure 2 shows the temporal profile of the laser pulse from the oscillator stage of the distributed-feedback dye laser. A single isolated pulse is obtained slightly above the threshold of laser oscillation. The pulse width becomes shorter with increasing pump energy. The shortest pulse (130 ps) is obtained at a pump energy that is 1.5 times above the threshold; however, additional peaks appear at higher pump energies. Thus, a short, single pulse can be obtained by optimizing the pulse energy of the pump laser. It should be noted that the laser pulse width can be changed by minor modification of the laser characteristics, such as beam diameter, divergence, and mono(19) Imasaka, T.; Proc. SPIE (Environ. Monit. Rem. Technol.) 1998, 3534, 573. (20) Onoda, T.; Saito, G.; Imasaka, T. Anal. Chim. Acta 2000, 412, 213-219.

Figure 3. Laser wavelength measured at different temperatures of the dye solution.

Figure 2. Temporal profile of a dye laser pulse at different pump energies. Pump energy/threshold pump energy (P/Pth): A, 1.1; B, 1.3; C, 1.5; D, 1.8; E, 2.0; F, 2.2; G, 2.5; H, 2.7; I, 3.0. The signal intensity is normalized by the intensity of the highest peak.

chromaticity, by changing the pump energy. This may be preferential for a quantitative discussion in a comparison of the ionization efficiencies obtained by the picosecond and nanosecond lasers. The spectral profile of the dye laser was measured by the wavemeter. The observed spectrum consisted of a single line whose line width was 0.01 nm. The time-bandwidth product is 0.39, which is close to the Fourier transform limit (0.441 and 0.315 for Gaussian and Sech2 pulses, respectively). As previously described, the drift of the laser wavelength represents a serious problem that needs to be solved for the practical application of a distributed-feedback dye laser to spectroscopy and spectrometry. Figure 3 shows the emitting wavelength of the dye laser at different dye solution temperatures. The shift in the laser wavelength is precisely identical to the value calculated using eqs 1 and 2. It should be noted that this behavior is sometimes used, even in wavelength tuning of the distributed-feedback dye laser.15,21 This slope suggests that the temperature drift should be maintained at