Chapter 3 Femtosecond time-resolved spectroscopy
3.2 Experimental setup
For the pump-probe measurements, a noncollinear optical parametric amplifier (NOPA) [37-39] was used as a light source as shown in Figure 3-4. The pump source of a NOPA system was a regenerative amplifier (Coherent, Legend L-USP-1K-HE) with the central wavelength of 800 nm, the pulse duration of 30 fs, the repetition rate of 5 kHz, and the average output power of 2.3 W. The output light of regenerative amplifier was split into two beams by a beam splitter. The fraction of output light was doubled in frequency by a BBO crystal and used as the pump for NOPA. On the other hand, the remainder light was focused into a sapphire plate to generate white-light continuum as the seed for NOPA. The output pulses from the NOPA were compressed with an optical pulse compressor composed of a diffraction grating telescopic dispersion line, specially designed multiplayer dielectric chirped mirrors, and a computer-controlled flexible mirror. After the compressor, the laser pulses were separated into two beams, one was used as a pump beam, the other was used as a probe beam after passing through an optical delay line. The setup was optimized to obtain pump and probe energies of 8 and 0.8 nJ, respectively. Both of pump and probe pulses were shorter than 10 fs
Figure 3-4 Schematic diagram of the visible sub-10 fs NOPA system and pump-probe experimental setup. PD: photodiode; BS: beam splitter; SP: sapphire plate; SM: spherical mirror; CM: ultrabroad-band chirped mirror; GR: 300-lines/mm ruled diffraction grating; QB:
10-cm quartz block; PS: periscope; VND: variable neutral-density filter.
in duration and covered the spectral range of 514-758 nm. Within the pulses, the spectral phase is almost constant resulting in nearly the Fourier-tansform limited pulses. The pump beam was mechanically chopped at 2.5 kHz in synchronization with the laser pulses at 5 KHz.
The pump and probe beams were set collinear and focused on the samples. Then the probe beam intensity was synchronously detected by a Si photodiode attached to a monochromator (Prinston Instruments, SpectraPro 2300i). The intensity changes of transmitted probe beam induced by the pump were processed through a 128 channel lock-in amplifier referenced to the frequency and phase of chopper in pump beam.
3.3 Experimental detail
In this study, we performed both the shorter time region and the longer time region pump-probe experiments. The delay-time ranges were set between -300 fs and 1400 fs and between -0.9 ps and 14.5 ps, respectively. There are 500 data point in time region and 96 data points in the spectral range between 514 nm and 758 nm in each experimental data, and step width was 3.6 fs in the shorter time region experiment and 31 fs in the longer time region experiment. For these experiments, the full width at half maximum (FWHM) of the pulse was determined by second-harmonic generation frequency-resolved optical gating (SHG-FROG) measurements [40, 41] to be slightly shorter than 10 fs, within which it carried a nearly constant spectral phase. The pulse energies of the pump and probe were about 8 nJ and 0.8 nJ, respectively.
During the measurements, a liquid DR19 sample was taken in a 1-mm-thick quartz flow cell (Starna, Type : 48-Q-1) , and rapidly recirculated with a peristaltic pump through a 4 mm i.d. Tygon tube (R3603) to ensure the fresh sample after each laser shot. The configuration of circulation system was shown in Figure 3-5 (a). Large volumes of the liquid samples with 60 ml were chosen to avoid accumulation of photo-generated cis isomer. For the experiment of polymer films, the samples were placed on the holder which was connected with a vertical stage as shown in Figure 3-5 (b). The measured position of a film was varied after every measurement to avoid the damage on the surface caused by laser pulses. In case of the polymer films, the effect of accumulation may be considerable (The cis isomer thermally reforms the trans isomer with a time constant of several seconds or milliseconds) [42]. All the experiments were performed at room temperature.
Figure 3-5 The photograph of the measurements for (a) a liquid sample and (b) a polymer film.
Chapter 4
Results and discussion
4.1 Femtosecond absorption spectroscopy in the shorter time region (from -300 fs to 1400 fs)
4.1.1 2D absorbance changes
The experimental data were represented with a two-dimensional (2D) plot of the transmittance changes (ΔT/T) as a function of both the probe frequency and delay time.
According to the calculating formula , one can obtain the absorbance changes. Figure 4-1 shows the two-dimensional view of absorbance change spectrum, which has 500 data points in time region between -300 fs and 1400 fs with the interval of 3.6 fs and 96 data points in spectral region between 514 nm and 758 nm. The contour line of ΔA = 0 is shown as black line. The photo induced absorbance changes ΔA for solution samples is negative in a spectral range of 514 nm to 578 nm and positive beyond this spectral range.
Additionaly, the negative ΔA for polymer films is in a spectral range of 514 nm to 591 nm.
The negative ΔA is caused by the absorption bleaching, which means the population of electrons depleted when DR19 was photoexcited by pump pulses. The positive ΔA is a result of the induced absorption of the excited state. The intense signal of ΔA near zero-delay time is attributed to a coherent spike, and the interference due to the superimposition of pump pulses and probe pulses is various at different wavelengths. Figure 4-2 and 4-3 show the ΔA as a function of delay time of DR19 in solution and a polymer film, respectively, for probe wavelengths of 558 nm, 583 nm, 609 nm, 635 nm, 660 nm, 686 nm, and 712 nm. The
560 580 600 620 640 660 680 700 720 18000 17000 16000 15000 14000
Probe frequency (cm-1)
(a)
560 580 600 620 640 660 680 700 720 -200 18000 17000 16000 15000 14000
Probe frequency (cm-1)
(b)
Figure 4-1 2D transient absorbance change spectrum of DR19 in (a) solution and (b) a polymer film.
-0.06 -0.04 -0.02 0.00
0.02 -200 0 200 400 600 800 1000 1200 1400
-0.04
-200 0 200 400 600 800 1000 1200 1400
-0.05
Figure 4-2 Transient absorbance changes of DR19 in solution at various probe wavelengths.
-0.04 -0.02 0.00 0.02
0.04 -200 0 200 400 600 800 1000 1200 1400
-0.01
-200 0 200 400 600 800 1000 1200 1400
-0.002
Figure 4-3 Transient absorbance changes of DR19 in a polymer film at various probe wavelengths.
negative photoinduced absorbance changes found at probe wavelength 558 nm are attributed to the bleaching of the ground state population. The absorbance changes is close to zero at 583 nm, where is neither bleaching nor photoinduced absorption. At the other probed wavelengths, the positive sign of ΔA is caused by the induced absorption of the excited state.
The photoinduced absorbance changes exist up to 1400 fs due to the lifetime longer than the measured time region. These signals are composed of a slow relaxation dynamics and a periodical oscillation. The former is assigned to the electronic relaxation, and the latter is attributed to molecular vibration. The oscillatory signals exist up to 1400 fs after excitation, and they seem to be composed of two dominate frequencies. The Fourier transform of the periodical oscillation will reveal information about the vibrational amplitude and phase.
4.1.2 Lifetime fitting
The time-dependent absorbance changes is analyzed by the double-exponential decay fitting. Figure 4-4 shows the lifetime at all probed wavelengths. In these results with short scanning range in delay time, the longer lifetime cannot be fitted well and it will be discussed in subsection 4.2.2. However, the fast decay with a time constant of ~100 fs was observed in both samples. It has been attributed to a large amplitude motion of the excited molecule out of the Franck-Condon region [43]. Additionally, this attribution of the rapid internal conversion from the S2 state to the S1 state is also proposed in previous works by other groups [44, 45].
These experimental results indicate that the fast dynamic behavior in solution samples is similar to that in the polymer films.
180000 17000 16000 15000 14000
Probe frequency (cm-1)
560 580 600 620 640 660 680 700 720 Probe wavelength (nm)
t1~103 fs
(a)
180000 17000 16000 15000 14000 50
Probe frequency (cm-1)
560 580 600 620 640 660 680 700 720 Probe wavelength (nm)
(b)
Figure 4-4 Lifetime of DR19 in (a) solution and (b) a polymer film.
4.1.3 Fast Fourier transform of the absorbance changes
With a multi-channel lock-in amplifier, the time-dependent absorbance changes were taken over the spectral range from 514 nm to 578 nm. The Fourier power spectra of oscillating components were obtained from the real-time trace by fast Fourier transform (integrated over the time range from 100 fs to 1400 fs since there are interference signals caused by the probe pulses and scattered pump pulses in the delay time earlier than 100 fs).
The two-dimension plots of FFT power as a function of the molecular vibrational frequencies and probe wavelengths are shown in Figure 4-5. Several vibrational modes appeare in FFT power spectrum, especially two prominent modes around 1105 cm-1 and 1320 cm-1. These modes may be expressed as the modes mixing due to the small difference in frequency between the adjacent modes, assigned to the stretching of C-C and symmetric stretching of NO2, respectively. Both signals are also observed in the stationary Raman spectrum, e.g.
Figure 2-5.
Figure 4-6 and 4-7 show the Fourier amplitude spectrum of real-time traces for probe wavelengths of 555 nm, 573 nm, 591 nm, 609 nm, 627 nm, 645 nm, 663 nm, and 681 nm. The intense peaks in the spectrum of a solution sample located around 470 cm-1, 1105 cm-1, 1180 cm-1, 1320 cm-1, and 1560 cm-1 which are also observed in a polymer film (shown in Fig. 4-7) except for the 470 cm-1 attributed to the solvent molecule of TMTE. In the FFT amplitude spectra, the dominant modes at 1105 cm-1 and 1320 cm-1 is strong probe wavelength dependence, and the FFT amplitude of peak in 1105 cm-1 is drastically reduced at 591 nm in the solution case. The peak at 1180 cm-1 assigned to the C-C stretching and coupled to C-N stretching is more intense in the solid state. For the pump-probe measurements, solutions of DR19 were recirculated with a peristaltic pump, and the pump-driven flow system was used to ensure that the laser pulses encountered fresh sample. However, the pulses were focused on the fixed point of a film, and the samples may have been damaged during the pump-probe
560 580 600 620 640 660 680 700 720 18000 17000 16000 15000 14000
Probe frequency (cm-1)
(a)
560 580 600 620 640 660 680 700 720 0 18000 17000 16000 15000 14000
Probe frequency (cm-1)
(b)
Figure 4-5 FFT power spectrum of vibration in the absorbance changes. DR19 in (a) solution and (b) a polymer film.
0 5 10 15
0 200 400 600 800 1000 1200 1400 1600 1800 2000
0
0 200 400 600 800 1000 1200 1400 1600 1800 2000
0
Figure 4-6 FFT amplitude spectra of DR19 in solution.
0 2 4 6
0 200 400 600 800 1000 1200 1400 1600 1800 2000
0.0
0 200 400 600 800 1000 1200 1400 1600 1800 2000
0.0
Figure 4-7 FFT amplitude spectra of DR19 in a polymer film.
measurements. For this reason, the absorbance changes in solution samples are larger than that in the polymer films, and the corresponding Fourier amplitudes of molecular vibrations in solution is also larger than that in the polymer films.
4.1.4 Analysis of molecular vibrations
In order to investigate the mode frequencies and the phase of the wavepacket motion, the probe wavelength-dependent oscillation at ~1105 cm-1 and ~1320 cm-1 were analyzed by the integration of FFT power with the band width 100 cm-1, tracking the peak position, and the phase variations of the wavepacket motion in whole spectral range. The analytic results of 1105 cm-1 and 1320 cm-1 are shown in Figure 4-8 and 4-9, respectively. The FFT power of the C-C stretching (1105 cm-1) clearly exhibit intense band in the spectral range of 550 nm to 580
18000 17000 16000 15000 14000 -1
0 1 FFT PowerPeak position (cm-1 )Phase (π)
Probe frequency (cm-1)
560 580 600 620 640 660 680 700 720 Probe wavelength (nm)
18000 17000 16000 15000 14000 -1
0 1 FFT PowerPeak position (cm-1 )Phase (π)
Probe frequency (cm-1)
560 580 600 620 640 660 680 700 720 Probe wavelength (nm)
(b)
Figure 4-8 Reproducible peak tracking of molecular vibration at 1105 cm-1 in (a) solution and (b) a polymer film.
0
18000 17000 16000 15000 14000 -1
0 1 FFT PowerPeak position (cm-1 )Phase (π)
Probe frequency (cm-1)
560 580 600 620 640 660 680 700 720
(a)
18000 17000 16000 15000 14000 -1
0 1 FFT PowerPeak position (cm-1 )Phase (π)
Probe frequency (cm-1)
560 580 600 620 640 660 680 700 720
(b)
Probe wavelength (nm)
Figure 4-9 Reproducible peak tracking of molecular vibration at 1320 cm-1 in (a) solution and (b) a polymer film.
nm and 615 nm to 645 nm. The peak position of the C-C stretching almost keep constant with 1105 cm-1 in whole spectral range, as shown in Figure 4-8. Additionally, the phase is around zero, -π or π in both samples, which corresponds to a wavepacket motion in the excited state.
The probe wavelength-dependent FFT power of the NO2 symmetric stretching (1320 cm-1) shown in Figure 4-9 are similar to the FFT power of the C-C stretching (1105 cm-1). The peak position of the NO2 symmetric stretching are around 1320 cm-1 for both samples in solution and a polymer film. However, the phase of wavepacket motion seems be modulated by the frequency of ~900 cm-1 in solution samples, which is absent in a polymer film.
4.2 Femtosecond absorption spectroscopy in the longer time region (from -0.9 ps to 14.5 ps)
4.2.1 2D absorbance changes
Since the isomerization proceeds of azo dye with a time constant is greater than 2 ps, it is hard to determine the electronic relaxation time of isomerization of azo dye within the shorter delay-time range. So the pump-probe experiments with longer delay time were also performed in this study. Figure 4-10 shows the two-dimensional view of absorbance change spectrum, which has 500 data points in time region between -0.9 ps and 14.5 ps with the interval of 31 fs and 96 data points in spectral region between 514 nm and 758 nm. Black lines represent the contour line of ΔA = 0. In the longer wavelength range (from 640 nm to 720 nm), the intense absorbance changes of DR19 in solution is positive and decay within 3 ps. However, in the spectral range between 590 nm and 640 nm, the decay is not clearly observed and the relaxtion dynamics is different from that in longer wavelength range from 640 nm to 720 nm. Moreover, this significant characteristics of DR19 in solution are obviously smeared out in the spectrum of DR19 in a polymer film as shown in Fig. 4-10(b).
Figure 4-11 and 4-12 show the transient absorbance changes at various probe wavelengths. Due to the scanning step with 31 fs is larger than the oscillation periods of molecular vibrations, the periodical component is absent in the transient absorbance change spectrum. The positive photoinduced absorbance changes are clearly observed in both cases of solution and a polymer film. However, the negative ΔA only appear arround 558 nm in a polymer film. The further analysis of the transient absorbance changes will be shown in next two subsections.
580 600 620 640 660 680 700 720
17000 Probe frequency (cm16000 15000-1) 14000
580 600 620 640 660 680 700 720 0
Probe frequency (cm-1)
(b)
Figure 4-10 2D transient absorbance change spectrum of DR19 in (a) solution and (b) a polymer film.
-0.02 Figure 4-11 Transient absorbance changes of DR19 in solution at various probe wavelengths.
-0.010
Figure 4-12 Transient absorbance changes of DR19 in a polymer film at various probe wavelengths.
4.2.2 Time-resolved spectrum
The transient absorption spectra at various delay time are shown in Figure 4-13 and 4-14. In order to obtain the smoother curve, the data points adjoined delays (±0.5 ps) were averaged to be one data point of absorbance change in each delay time. As shown in Figure 4-13, in the range of longer wavelengths between 640 nm and 710 nm, the absorbance changes decrease dramatically in the delay time earlier than 2 ps, and decrease slowly in the later delay times. The rapid decay is caused by the photoinduced absorption in the excited state, but the decay is slow in the spectral range between 595 nm and 640 nm. In the range of shorter wavelengths between 555 nm and 580 nm, the ΔA increased with time and the bleaching band does not return to zero during the measurements. The transient absorption spectra of DR19 in a polymer film are different from the results in solution, as shown in Figure 4-14. The absorbance changes decrease with similar rate in the range of wavelengths between 620 nm and 700 nm.
17500 17000 16500 16000 15500 15000 14500 14000 -2
Probe frequency (cm-1)
580 600 620 640 660 680 700 720 Probe wavelength (nm)
Figure 4-13 Transient absorption spectra at various delay time for DR19 in solution.
17500 17000 16500 16000 15500 15000 14500 14000 -1.5
Probe frequency (cm-1)
1 ps
580 600 620 640 660 680 700 720 Probe wavelength (nm)
Figure 4-14 Transient absorption spectra at various delay time for DR19 in a polymer film.
4.2.3 Electronic relaxation time
The transient absorbance changes in longer time region were also analyzed by the double-exponential decay fitting. Figure 4-15(a) and (b) show the lifetime of DR19 in solution and a polymer film at all probed wavelengths, respectively. The relaxation time of
~0.63 ps and ~2.55 ps in solution between 640 nm and 700 nm were attributed to the internal conversion to the ground state and vibrational cooling in the ground state, respectively, which is consistent with the studies of Schmidt et al. in a push-pull substituted azobenzene (4-nitro-4’-(dimethylamino)azobenzene, NA) [43]. In solution as shown in Figure 4-15(a), the shorter time constant keeps in the same value ~ 0.63 ps in most spectral range while the longer time constant is wavelength-dependent in the range between 590 nm and 640 nm.
For the case of DR19 in a polymer film as shown in Figure 4-15(b), both time
180000 17000 16000 15000 14000
Probe frequency (cm-1)
560 580 600 620 640 660 680 700 720 Probe wavelength (nm)
(a)
180000 17000 16000 15000 14000
1
Probe frequency (cm-1)
(b)
560 580 600 620 640 660 680 700 720
0.84 ps Probe wavelength (nm)
4.43 ps
Figure 4-15 Lifetime of DR19 in (a) solution and (b) a polymer film.
constants (~0.84 ps and ~4.43 ps) seem to be wavelength–independent between 590 nm and 640 nm. All of the relaxation processes in the excited state vanished within 2.55 ps in solution and 4.33 ps in a polymer film. This implies that the isomerization dynamics of trans-DR19 are sensitive to the molecular environment. The relaxation process of trans-DR19 in solution and a polymer film is sketched in Figure 4-16. Namely, the trans-DR19 will be excited from S0
ground state to S2 excited state (Franck-Condon state), and then relax to S1 state within ~100 fs (the black arrows in Figure 4-16). Due to the internal conversion, the trans-DR19 will relax to cis-DR19 or trans-DR19 with sub-ps (the red arrows in Figure 4-16). Finally, the cis-DR19 and trans-DR19 are, respectively, cooled from high-energy level to low-energy level of ground (S0) state within several ps (the blue arrows in Figure 4-16).
Figure 4-16 Schematic diagram of the relaxation of trans-DR19 after photoexcitation in solution and a polymer film.
Chapter 5 Summary
The photoisomerization of DR19 in solution and a polymer film were investigated by sub-10 fs time-resolved absorption measurements with broadband visible spectral range. The most important results are described as follows,
(1) In the shorter time range, the motion of a wavepacket in the excited state out of the Franck-Condon state occurs on the time scale of ~100 fs in both samples with the form of solutions and films. The vibrational frequencies obtained from the FFT power spectra of solution samples are consistent with the vibrational frequencies in polymer films, but there are slightly difference in the wavepacket motion associated to the symmetric stretching of NO2. These results indicate the molecular vibrations are not deeply influenced by the surrounding environment.
(2) In the longer time range, the internal conversion to the ground state processed with a time constant of ~0.63 ps in solution and ~ 0.84 ps in a polymer film. In addition, the longer time constant of ~2.55 ps in solution and ~4.43 ps in a polymer film are attributed to the vibrational cooling in the ground state.
These results suggest that the isomerization of DR19 is strongly influenced by the molecular environment. Furthermore, this study provides the essential information for photonics applications such as optical switches and optical memory storage devices, which put the photoisomerization mechanism in use.
References
[1] Hartley, G. S. Nature 1937, 281; J. Chem. Soc. 1938, 633-642 [2] Rau, H.; Lüddecke, E. J. Am. Chem. Soc. 1982, 104, 1616
[3] Siampiringue, N.; Guyot, G.; Monti, S.; Bortolus, P. J. Photochem. 1987, 37, 185 [4] Okamoto, H.; Hamaguchi, H.; Tasumi, M. Chem. Phys. Lett. 1986, 130, 185
[5] Biancalana, A.; Campani, E.; Gorini, G.; Masetti, G.; Quaglia, M. J. Raman Spectrosc.
1992, 23, 155
[6] Stuart, C. M.;Frontiera, R. R.; Mathies, R. A. J. Phys. Chem. A 2007, 111, 12072-12080 [7] Curtis, R. D.; Hilborn, J. W.; Wu, G.; Lumsden, M. D.; Wasylishen, R. E.; Pincock, J. A. J.
Phys. Chem., Int. Ed. Engl. 1993, 97, 1856
[8] Armstrong, D. R.; Clarkson, J.; Smith, W. E. J. Phys. Chem. 1995, 99, 17825
[9] Cembran, A.; Bernarli, F.; Garavelli, M.; Gagliardi, L.; Orlandi, G. J. Am. Chem. Soc.
2004, 126, 3234
[10] Conti, I.; Garavelli, M.; Orlandi, G. J. Am. Chem. Soc. 2008, 130, 5216
[11] Lednev, I. K.; Ye, T. Q.; Matousek, P.; Towrie, M.; Foggi, P.; Neuwhl, F. V. R.; Umapathy, S.; Hester, R. E.; Moore, J. N. Chem.Phys. Lett, 1998, 290, 68
[12] Bortollus, P.; Monti, S. J. Phys. Chem. 1979, 83, 648
[13] Rau, H. In Photochromism: Molecular and Systems; Dürr, H., Bouas-Lauran, H., Eds.;
Elservier: Amsterdarn, 2003; p 165
[14] Bortolus, P.; Monti, S. J. Phys. Chem. 1979, 83,648
[15] Wildes, P. D.; Pacifici, J. G.; Gether, I. J.; Whitten, D. G. J. Am. Chem. Soc. 1971, 93, 2004
[16] Liu, Z. F.; Hashimoto, K.; Fujishima, A. Nature 1990, 347, 658 [17] Sekkat, Z.; Dumont, M. Appl. Phys. B 1992, 54, 486
[18] Ikeda, T.; Tsutsumi, O. Science 1995, 268, 1873.
[19] Tamai, N.; Miyasaka, H. Chemical Reviews 2000, 100, 1875-1890
[20] Jung, C. C.; rutloh, M.; Stumpe, J. J. Phys. Chem. B 2005, 109, 7865-7871 [21] Muraoka, T.; Kinbara, K.; Aida, T. Nature 2006, 440, 512-515
[22] Fujino, T.; Arzhantsev, S. Yu.; Tahara, T. J. Phys. Chem. A 2001, 105, 8123 [23] Baltuska, A.; Fuji, T.; Kobayashi, T.; Opt. Lett. 2002, 27, 306
[24] Chang, C. W.; Lu, Y.-C.; Wang, T.-T.; Diau, E. W.-G. J. Am. Chem. Soc. 2004, 126, 3234 [25] Myers, A. B.; Mathies, R. A. J. Chem. Phys. 1984, 81, 1552
[26] Sension, R. J.; Repinec, S. T.; Szarka, A. Z.; Hochstrasser, R. M. J. Chem. Phys. 1993, 98, 6291 and references therein
[27] Polanyi, J. C.; Zewail, A. H. Acc. Chem. Res. 1995, 28, 119
[28] Lednev, I. K.; Ye, T. Q.; Hester, R. E.; Moore, J. N. J. Phys. Chem. 1996, 100, 13338 [29] Asano, T.; Okada, T.; Shinkai, S.; Shigematsu, K.; Kusano, Y.; Manabe, O. J. Am. Chem.
Soc. 1981, 103, 5161
[30] Shin, D. M.; Whitten. D. G. J. Am. Chem. Soc. 1988, 110, 5206 [31] Crecca, C. R.; Roitberg, A. E. J. Phys. Chem. A 2006, 110, 8188 [32] Biswas, N.; Umapathy, S. J. Phys. Chem. A 2000, 104, 2734
[33] Ando, R. A.; Rodriguez-Redondo, J. L.; Sastre-Santos, A.; Fernandez-Lazaro, F.;
Azzellini, G. C.; Borin, A. C.; Santos, P. S. J. Phys. Chem. A 2007, 111, 13452 [34] Clarkson, J.; Smith, W. E. J. Mol. Struct. 2003, 655, 413
[35] Biswas, N.; Umapathy, S. J. Raman Spectrosc. 2001, 32, 471
[36] Cerullo, G.; Manzoni, C.; Lüer, L; Polli, D. Photochem. Photobiol. Sci. 2007, 6, 135 [37] Shirakawa, A.; Sakane, I.; Kobayashi, T. Opt. Lett. 1998, 23, 1292
[38] Shirakawa, A.; Sakane, I.; Takasaka, M.; Kobayashi, T. Appl. Phys. Lett. 1999, 74, 2268
[38] Shirakawa, A.; Sakane, I.; Takasaka, M.; Kobayashi, T. Appl. Phys. Lett. 1999, 74, 2268