Dependence of the Laser Two-Photon Ionization Process in Solution

an IR beam. When the laser pulse width was 300 ps, it proceeded through a geminate pair state (a stepwise process), as indicated by the signal enhance...
0 downloads 0 Views 54KB Size
Anal. Chem. 2001, 73, 2066-2069

Dependence of the Laser Two-Photon Ionization Process in Solution on the Laser Pulse Width Teiichiro Ogawa,* Masahiro Mizutani,† and Takanori Inoue‡

Department of Molecular Science and Technology, Kyushu University, Kasuga-shi, Fukuoka 816-8580, Japan

The laser two-photon ionization process has been investigated using simultaneous irradiation of an UV beam and an IR beam. When the laser pulse width was 300 ps, it proceeded through a geminate pair state (a stepwise process), as indicated by the signal enhancement with simultaneous irradiation of the two laser beams. Although no signal enhancement was observed when the laser pulse width was 100 fs, and because molecules with no absorption at the laser wavelength showed an intense signal, the two-photon ionization excited by a femtosecond laser should proceed through a simultaneous two-photon process. The detection limit was quite susceptible to the laser fluctuations. Laser two-photon ionization is a useful technique for the highly sensitive determination of many aromatic and other photoabsorbing molecules in solution1-8 and on a surface.2,3,9-13 There have been a few investigations of the ionization efficiency in hydrocarbon solvents and its dependence on the wavelength and solvent properties.14-16 No investigation has, however, been carried out on the laser pulse-width dependence of the ionization process and the detection efficiency, although this information is required for * To whom correspondence should be addressed. Present address: Department of Molecular and Material Sciences, Kyushu University, Fukuoka 816-8580, Japan. † Present address: Tokyo Center for Quality Control & Consumer Service, Ministry of Agriculture, Forestry and Fisheries, Saitama 330-9731, Japan. ‡ Present address: Department of Applied Chemistry, Oita University, Oita 870-1192, Japan. (1) Ogawa, T. Prog. Anal. Spectrosc. 1986, 9, 429. (2) Ogawa, T. Handbook of Advanced Materials Testing; Cheremisinoff, N. P., Cheremisinoff, P. N., Eds.; Marcel Dekker: New York, 1995; p 27. (3) Ogawa, T. Photoionization and Photodetachment; Ng, C.-Y., Ed.; World Scientific: River Edge, NJ; 2000; p 601. (4) Voigtman, E.; Jurgensen, A.; Winefordner, J. D. Anal. Chem. 1981, 53, 1921. (5) Voigtman, E.; Winefordner, J. D. Anal. Chem. 1992, 54, 1834. (6) Yamada, S.; Kano, K.; Ogawa, T. Bunseki Kagaku 1982, 31, E247. (7) Yamada, S.; Hino, A.; Kano, K.; Ogawa, T. Anal. Chem. 1983, 55, 1914. (8) Yamada, S.; Sato, N.; Kawazumi, H.; Ogawa, T. Anal. Chem. 1987, 59, 2719. (9) Masuda, K.; Inoue, T.; Yasuda, T.; Nakashima, K.; Ogawa, T. Anal. Sci. 1993, 9, 297. (10) Inoue, T.; Masuda, K.; Nakashima, K.; Ogawa, T. Anal. Chem. 1994, 66, 1012. (11) Ogawa, T.; Yasuda, T.; Kawazumi, H. Anal. Chem. 1992, 64, 2615. (12) Gridin, V. V.; Korol, A.; Bulatov, V.; Schechter, I. Anal. Chem. 1996, 68, 3359. (13) Gridin, V. V.; Litani-Barzilai, I.; Kadosh, M.; Schechter, I. Anal. Chem. 1998, 70, 2685. (14) Ogawa, T.; Kise, M.; Yasuda, T.; Kawazumi, H.; Yamada, S. Anal. Chem. 1992, 64, 1217. (15) Ogawa, T.; Sato, M.; Tachibana, M.; Ideta, K.; Inoue, T.; Nakashima, K. Anal. Chim. Acta 1995, 299, 355. (16) Ogawa, T.; Tenkyuu, Y.; Nakashima, K. Anal. Sci. 2000, 16, 913.

2066 Analytical Chemistry, Vol. 73, No. 9, May 1, 2001

further improvements in highly sensitive detection based on twophoton ionization. Highly sensitive determinations based on this method published so far were carried out using a nanosecond laser, such as a nitrogen laser or a Nd:YAG laser. The two-photon ionization signal was roughly proportional to the absorbance at the laser wavelength.1-3,7 This finding indicates that the two-photon ionization process excited by a nanosecond laser proceeds through a real excited state.1-3 The two-photon ionization signal increased upon the simultaneous irradiation of the near-infrared beam when excited by a nanosecond Nd:YAG laser.17-19 This finding indicates that the two-photon ionization process excited by the nanosecond laser proceeds through a geminate ion-electron pair.2,3,17-21 However, there is no information about the signal dependence on the absorbance and signal enhancements when excited by a shorter-pulse-width laser. When the laser pulse width is much shorter than that of the intermediate states (a real excited state or a geminate pair), the ionization behavior may change. In this study, we carried out the two-photon ionization of several aromatic molecules using a picosecond Nd:YAG laser (300 ps) and a femtosecond Ti-sapphire laser (100 fs). The ionization process was investigated on the basis of the signal enhancement by simultaneous irradiation of the long-wavelength pulses. The sensitivity and detection limits were also compared to those of excitation using a nanosecond laser. EXPERIMENTAL SECTION The apparatus was constructed on the basis of the previous one,15,22 paying attention to the difference in the laser pulse width. The third harmonic (355 nm, 3.5 eV) of a picosecond Nd:YAG laser (Quantel, 15 mJ at 1064 nm, 300 ps) was used for the excitation source in the picosecond experiments. The maximum pulse energy of the third harmonic was about 1.0 mJ/pulse. The excitation beam was softly focused in a region using a lens (focal length, 90 cm) between two copper electrodes in a standard quartz fluorescence cell (10 × 10 × 40 mm).15 The irradiation area was approximately 0.02 cm2, and the typical laser energy was about 0.1 GW/cm2 for a 0.3 mJ/pulse. One of the electrodes was connected to a high-voltage supply unit, and the typical applied (17) Yamada, S. Anal. Chem. 1989, 61, 612. (18) Yamada, S.; Shino, I. Talanta 1989, 36, 937. (19) Yamada, S. Anal. Sci. 1991, 7, 223. (20) Holroyd, R. A.; Preses, J. M.; Zenos, N. J. Chem. Phys. 1983, 79, 483. (21) Holroyd, R. A.; Preses, J. M.; Bottcher, E. H.; Schmidt, W. F. J. Phys. Chem. 1984, 88, 744. (22) Nakashima, K.; Kise, M.; Ogawa, T.; Kawazumi, H.; Yamada, S. Chem. Phys. Lett. 1994, 231, 81. 10.1021/ac001055m CCC: $20.00

© 2001 American Chemical Society Published on Web 03/20/2001

Figure 2. Dependence of photoionization current on the laser pulse energy during the 300-ps excitation. Sample, naphthacene (1 × 10-6 M) in hexane.

Figure 1. Optics for a two-color femtosecond excitation. The amplified femtosecond laser beam (800 nm, 1 mJ) was frequencydoubled by a BBO crystal and separated using a dichroic mirror. Timing of the irradiation was adjusted using an optical delay for one of the beams.

voltage was 2.5 kV/cm. Another electrode was connected to a preamplifier (Keithley 428). The photoionization current was accumulated for 64 laser pulses using a digital storagescope (Sony Tektronix TDS620, 500 MHz), and the average of 8 independent measurements was taken as the signal. A Ti-sapphire laser (Coherent MIRA 900D, 200-fs) was used for the excitation in the femtosecond experiments. The femtosecond laser beam was amplified by a regenerative amplifier for a series of pulses with a pulse energy of 1 mJ/pulse (800 nm), a pulse width of 100 fs, and a repetition rate of 10 Hz. The second harmonic (400 nm, 3.1 eV) of this pulse (maximum, 0.2 mJ/pulse) generated by a BBO crystal was softly focused using a concave mirror (focal length, 40 cm) in a standard quartz fluorescence cell (same as above). The irradiation area was approximately 0.08 cm2, and the typical laser energy was about 10 GW/cm2, for a 0.1 mJ/pulse. A flat capillary cell, which was 6 mm in width and 0.3 mm in depth, was used for the highly sensitive detection with the femtosecond laser, and two platinum wires (0.1 mm in diameter) served as the two electrodes in this cell. The experimental apparatus for the simultaneous irradiation of both the 400and 800-nm laser beams is shown in Figure 1. Timing of the two laser beams was adjusted by changing the length of the optical delay path and was measured by an auto-correlator. The electronics were the same as above. The signal was accumulated for 32 laser pulses, and the average of 8 independent measurements was taken as the signal. Pyrene (Janssen Chemica, 99%) and other aromatic molecules were used as supplied. The solvents used were reagent grade and were used after distillation. RESULTS AND DISCUSSION Picosecond and Femtosecond Excitations. The photoionization signal was quadratically proportional to the laser pulse energy, as shown for naphthacene during picosecond excitation in Figure 2. This finding indicates that the excitation is a twophoton process, as in the case of the nanosecond excitation. The photoionization signals of pyrene, anthracene, and perylene in isooctane were proportional to the concentration, as typically shown for pyrene in Figure 3. The photoionization signal of naphthacene in hexane during femtosecond excitation at 400 nm was quadratically proportional

Figure 3. Dependence of photoionization current on concentration during the 300-ps excitation. Sample, pyrene in isooctane; applied voltage, 0.5 kV/2 mm.

Figure 4. Dependence of photoionization current on the laser pulse energy during the 100-fs excitation. Samples: 4, naphthacene (1 × 10-5 M) in hexane; O, hexane.

to the laser pulse energy, and that of hexane was cubicly proportional to it, as shown in Figure 4. Thus, the excitation and ionization of naphthacene was a two-photon process, but that of the solvent was a three-photon process, as in the case of the nanosecond and picosecond excitations. There are two photocurrent thresholds of pyrene in hexane: one at 5.87 eV and the other at 6.34 eV.23 Thus, the two-photon process is reasonable. The photoionization signal would be larger at a higher applied voltage, because a higher applied field would increase the ionization efficiency.5,6 The signal-to-blank ratio increased at the higher applied voltage as expected and is shown in Figure 5 for naphththacene; thus, an applied field of 222 V/mm was used thereafter for the femtosecond excitation. The two-photon ionization signals of anthracene and isooctane during the picosecond and femtosecond excitations were compared to those during the nanosecond excitation in Table 1. The photoionization current of anthracene seems to be smaller during the picosecond excitation, but it was approximately identical to (23) Ogawa, T.; Murata, D.; Soga, H.; Nakashima, K. Bull. Chem. Soc. Jpn. 2001, in press.

Analytical Chemistry, Vol. 73, No. 9, May 1, 2001

2067

Figure 5. Dependence of photoionization current on the applied voltage. Sample: 4, naphthacene (1 × 10-6 M) in hexane; O, hexane. The laser pulse width was 100 fs. Table 1. Photoionization Signal of Anthracene (1 × 10-7 Mol/dm3) in Hexane

excitation source nanosecond laser (355 nm) picosecond laser (355 nm) femtosecond laser (400 nm)

laser pulse solvent photoionization energy, blank, current of anthracene, mJ/pulse nA nA 1 0.33 0.1

0.0 0.7 8.2

25 2.3 28

that during the nanosecond excitation, because the photocurrent was quadratically proportional to the laser pulse energy, and there was a difference of a factor of 3 in the pulse energy. It is noteworthy that anthracene, which has no absorption at the laser wavelength (400 nm), shows an intense photoionization signal during the femtosecond excitation. This finding indicates that the ionization is a simultaneous two-photon process when excited by a femtosecond laser. The solvent blank for the picosecond and femtosecond excitations was much larger than that for the nanosecond excitation, probably because the larger peak power of the picosecond and femtosecond beams induced a simultaneous three-photon ionization of the solvent molecules. Amplification Using Simultaneous Irradiation. The laser two-photon ionization under nanosecond excitation proceeds through an intermediate state called a geminate ion-electron pair in which the ionized electron and the produced ion still stay in the original solvent cavity.1,3,20,21,24,25 Most of the geminate pairs recombine and create no substantial photocurrent. If the separation efficiency of the geminate pairs can be increased, the photocurrent would substantially increase. Because the geminate pair has a broad absorption band in the near-infrared region, its ionization would be enhanced upon irradiation of a near-infrared laser.17-19 Furthermore, if the signal increases upon the irradiation of the near-infrared laser beam, we can conclude that the ionization proceeds through the geminate-pair. The infrared beam (fundamental, 1064 nm) was simultaneously introduced in the cell with the UV beam (third harmonic, 355 nm) during the picosecond excitation. No photocurrent was observed by irradiation of only the infrared beam. The pulse train of a 300ps pulse was about 9 cm in length, and the simultaneous irradiation of the two beams was monitored using a photodiode and oscilloscope. The photoionization signals of aromatic molecules, which have an absorption band at 355 nm, were amplified with the simultaneous irradiation. The amplification factor (ratio of the signals excited by 355 nm + 1064 nm to that excited only by 355 nm) in the cases of anthracene and 1-aminopyrene was 2.0 and 1.8, respectively, during the picosecond excitation and 2068 Analytical Chemistry, Vol. 73, No. 9, May 1, 2001

Figure 6. Dependence of photoionization current on concentration during the 100-fs excitation. Sample, naphthacene in hexane; applied voltage, 222 V/mm.

2.5 and 2.1,19 respectively, during the nanosecond excitation. There was no substantial difference in amplification between the nanosecond excitation and picosecond excitation. Thus, the two-photon ionization was a stepwise process through the geminate pair in both cases. The reason for the amplification was, as discussed by Yamada, due to the ionization of the geminate pairs.17-19 The lifetime of the geminate pair was estimated to be a few picoseconds,24,25 and the laser pulse width of 5 ns to 300 ps seems to induce no major difference in the amplification factor. A simultaneous irradiation of a femtosecond pulse beam of 400 nm and that of 800 nm was carried out using the optics shown in Figure 1. No photoionization signal was observed by irradiation of only the 800 nm beam. The photoionization signals of anthracene, naphthacene, pyrene, and benzo[a]pyrene were quadratically proportional to the pulse energy of the 400-nm beam and did not depend on that of the 800 beam, as typically shown for naphthacene in Figure 6. Contrary to the nanosecond and picosecond excitation, no amplification was observed; this finding may be due to the very short laser pulse width, as compared to the geminate-pair lifetime. Although anthracene, pyrene, and benzo[a]pyrene have no absorption band at 400 nm, and naphthacene does, there was no major difference in the photoionization signal and its amplification. These findings indicate that the twophoton ionization excited by a femtosecond pulse proceeds through a simultaneous two-photon process without passing through the real excited state and the geminate-pair state. This is because the peak energy of the femtosecond laser pulse is very strong, and a simultaneous process can proceed at a higher efficiency than a stepwise process, even if the molecule has an absorption band at the laser wavelength. Application to Highly Sensitive Determination. The picosecond and femtosecond lasers were not so stable as a typical nanosecond laser, and its power fluctuation created a large source of noise. A typical fluctuation was 0.082 ( 0.004 mJ/pulse for the nanosecond laser (1 mJ/pulse), 0.157 ( 0.052 mJ/pulse for the picosecond laser (0.33 mJ/pulse) and 0.058 ( 0.031 mJ/pulse for the femtosecond laser (0.1 mJ/pulse). The picosecond and femtosecond lasers are less stable than the nanosecond laser. Due to the large solvent blank during the picosecond and femtosecond excitations as a result of the three-photon process, a smaller laser pulse energy would be more preferable for a highly sensitive determination, but electrical noise would become more dominant at the smaller signals. A compromise is unavoidable. (24) Yakovlev, B. S.; Lukin, L. V. Photodissociation and Photoionization; Lawley, K. P., Ed.; John Wiley & Sons: New York, 1985; p 99. (25) Warman, J. M. The Study of Fast Processes and Transient Species by Electron Pulse Radiolysis; Baxendale, J. H., Busi, F., Eds.; D. Reidel: Dordrecht, The Netherlands, 1982; 433.

Table 2. Detection Limits (pg/mL) of Aromatic Molecules laser excitation source anthracene pyrene naphthacene perylene a

nanosecond, 0.3-0.4 mJ

picosecond, 1 mJ

10a 6a 60b 60b

500 40

femtosecond, 0.1 mJ

18 112

Ref 6. b Ref 7. Figure 7. Dependence of photoionization charge on the laser pulse energy during the 100-fs excitation. Sample, naphthacene (1 × 10-7 M) in hexane. Irradiation, O, 400 nm only; 4, 400 nm + 800 nm.

The detection limit of anthracene during the picosecond excitation was 0.5 ng/mL, and that6 for the nanosecond excitation was 10 pg/mL, as indicated in Table 2. This difference comes from the large solvent blank and from the laser power fluctuation in the present experiments. The detection limit during the femtosecond excitation was rather poor when a standard cell was used for the photoionization, because a thick solvent layer produced a femtosecond white light and the white light created photoelectrons from the electrode surface. Thus, it would be desirable to reduce the thickness of the cell and the surface area of the electrodes. A flat capillary cell was then constructed to overcome this problem. Its small depth (0.3 mm) and two thin wire electrodes (0.1 mm) were useful in reducing the effect of the femtosecond white light. A typical analytical curve of naphthacene obtained in this cell is shown in Figure 7. The detection limits during the femtosecond excitation were as small as those during the nanosecond excitation, as also indicated in Table 2. Rather favorable detection limits using the femtosecond laser came from improvements in the cell design.

Concluding Remarks. The two-photon ionization process depended on the laser pulse width. It proceeds through a stepwise process for 10 0.3-ns laser pulses but through a simultaneous process for a 100 fs pulse. The former process can be enhanced by the simultaneous irradiation of the infrared beam, which indicates a contribution of the geminate pair state, while the latter cannot. No enhancement was observed for the femtosecond laser excitation. The detection limit was rather poor using the picosecond laser. Stability of the laser and the suppression of the three-photon ionization signal of the solvent molecules are essential for improving the detection limit.

Received for review September 5, 2000. Accepted January 29, 2001. AC001055M

Analytical Chemistry, Vol. 73, No. 9, May 1, 2001

2069