**2. X-ray pinpoint structural measurement system for investigation of optical recording process**

The pulse characteristic and high coherent X-ray beam of SPring-8 allow us to investigate dynamics of chemical reactions and phase transition of materials caused by applied field. In order to realize the direct investigation, we developed the X-ray pinpoint structural measurement system at BL40XU of SPring-8, which enables the advanced X-ray measurement technique in nanometer spatial scale and/or picosecond time scale. This system was used not only for investigation of optical recording process but also for charge density level visualization of photo-induced phase transition (Fukuyama et al., 2010) and sub-micron single crystal structure analysis (Yasuda et al., 2008; Yasuda et al., 2010).

In order to investigate the amorphous-crystal phase change on recording DVD media we optimized the X-ray pinpoint structural measurement system. Figure 2 shows the schematic illustration of experimental setup for taking a snapshot of X-ray diffraction profile of optical phase-change materials. The main optimized components are laser-SR timing control system, a sample rotation system, an optical reflectivity probe system, and an X-ray diffractometer. The following sub-sections describe the components and their performance.

Fig. 2. Experimental setup for taking a snapshot of X-ray diffraction profile of optical phasechange materials.

### **2.1 Laser-pump and X-ray SR probe system**

A pulsed nature of synchrotron X-ray beam due to the acceleration mechanism of charged particles has enabled us to conduct the stroboscopic X-ray measurement in picosecond time-

Time Resolved Investigation

length was ~40 ps (FWHM).

with a time delay of τ.

delay time.

of Fast Phase-Change Phenomena in Rewritable Optical Recording Media 263

storage ring was used for the measurement. The filling pattern was set to be hybrid mode, in which the high current electron bunch (3 mA) exists with lower current bunch train (see Fig. 5(a)). In the filling pattern, the pulse width of SR, which is determined by the electron bunch

Fig. 5. Time chart of the pulsed laser and X-ray SR irradiation for a snapshot measurement

The control technique of the timing delay, τ, is also a key to make a stroboscopic measurement. As the time-scale of completing the phase change is a microsecond, the large delay is required, which is generally difficult to obtain with an optical delay. Thus, we developed a high precision RF phase control delay circuit to scan the time delay without degrading the precision (Fukuyama et al., 2008b). Figure 6 shows the photograph of the RF phase control circuit, whose timing jitter between the output and the input RF signal is about 3 ps (in rms), which was evaluated with a 16 GHz digital oscilloscope (Tanaka et al., 2010). The timing control precision in the whole system was 8.4 ps (in rms), which was estimated by time-resolved diffraction measurement using a fast lattice response of semiconductor single crystal. Since the distribution of the crossing points is narrower than

the pulse width of the SR, the pump-probe system has a time resolution of < 40 ps.

Fig. 6. High precision RF phase control circuit. The circuit provides a delayed RF signal from the output port with a precision of better than 3 ps (in rms) which is independent of the

scale. In the stroboscopic X-ray diffraction measurement, the pulsed X-rays from a SR source are diffracted at a sample after a short-pulsed laser synchronized to the SR X-ray pulses illuminates it, as schematically shown in Fig. 3. The time-dependent snapshot can be taken by changing the time delay indicated as τ in Fig. 3.

Fig. 3. Schematic illustration of time-resolved X-ray diffraction measurement using a laserpump X-ray-probe method.

To measure an impulse response of the crystallization, a mode-locked Ti:sapphire laser with a pulse width, Δ tL, of 130 fs was used as a pump laser. A photo of the laser system is shown in Fig. 4. The laser system is equipped with a regenerative amplifier which produces pulses with a pulse energy of 1 mJ, a wavelength of 800 nm, and a repetition rate of 1 kHz. The laser was synchronized to the RF master oscillator of the SR storage ring for acceleration of electrons, as shown in Fig. 2. Since the detailed synchronization system was published (Tanaka et al., 2000), the synchronization and timing control system are briefly summarized here. Timing synchronization of the mode-locked laser oscillator was achieved by controlling the cavity length, with the precision of a few ps. The repetition rate is 84.76 MHz, corresponding to the 1/6 of RF of the storage ring master oscillator. Output timing of regeneratively amplified laser pulses is controlled in consideration of the revolving frequency (the round trip is ~5 μs) of the storage ring.

Fig. 4. Femtosecond pulsed laser system. The laser system is mainly composed of a Ti:sapphire laser oscillator and a regenerative amplifier. The regenerative amplifier produces optical pulses with a pulse energy of 1 mJ, a wavelength of 800 nm, and a repetition rate of 1 kHz.

Figure 5 shows a time chart of the laser and the SR pulses in the experiment. The time structure of SR pulses is dependent on the filling pattern of electron bunches in the storage ring. Because the repetition rate of the measurement was ~1 kHz, only one bunch in the

scale. In the stroboscopic X-ray diffraction measurement, the pulsed X-rays from a SR source are diffracted at a sample after a short-pulsed laser synchronized to the SR X-ray pulses illuminates it, as schematically shown in Fig. 3. The time-dependent snapshot can be taken

Fig. 3. Schematic illustration of time-resolved X-ray diffraction measurement using a laser-

To measure an impulse response of the crystallization, a mode-locked Ti:sapphire laser with a pulse width, Δ tL, of 130 fs was used as a pump laser. A photo of the laser system is shown in Fig. 4. The laser system is equipped with a regenerative amplifier which produces pulses with a pulse energy of 1 mJ, a wavelength of 800 nm, and a repetition rate of 1 kHz. The laser was synchronized to the RF master oscillator of the SR storage ring for acceleration of electrons, as shown in Fig. 2. Since the detailed synchronization system was published (Tanaka et al., 2000), the synchronization and timing control system are briefly summarized here. Timing synchronization of the mode-locked laser oscillator was achieved by controlling the cavity length, with the precision of a few ps. The repetition rate is 84.76 MHz, corresponding to the 1/6 of RF of the storage ring master oscillator. Output timing of regeneratively amplified laser pulses is controlled in consideration of the revolving

Fig. 4. Femtosecond pulsed laser system. The laser system is mainly composed of a Ti:sapphire laser oscillator and a regenerative amplifier. The regenerative amplifier produces optical pulses with a pulse energy of 1 mJ, a wavelength of 800 nm, and a

Figure 5 shows a time chart of the laser and the SR pulses in the experiment. The time structure of SR pulses is dependent on the filling pattern of electron bunches in the storage ring. Because the repetition rate of the measurement was ~1 kHz, only one bunch in the

by changing the time delay indicated as τ in Fig. 3.

frequency (the round trip is ~5 μs) of the storage ring.

pump X-ray-probe method.

repetition rate of 1 kHz.

storage ring was used for the measurement. The filling pattern was set to be hybrid mode, in which the high current electron bunch (3 mA) exists with lower current bunch train (see Fig. 5(a)). In the filling pattern, the pulse width of SR, which is determined by the electron bunch length was ~40 ps (FWHM).

Fig. 5. Time chart of the pulsed laser and X-ray SR irradiation for a snapshot measurement with a time delay of τ.

The control technique of the timing delay, τ, is also a key to make a stroboscopic measurement. As the time-scale of completing the phase change is a microsecond, the large delay is required, which is generally difficult to obtain with an optical delay. Thus, we developed a high precision RF phase control delay circuit to scan the time delay without degrading the precision (Fukuyama et al., 2008b). Figure 6 shows the photograph of the RF phase control circuit, whose timing jitter between the output and the input RF signal is about 3 ps (in rms), which was evaluated with a 16 GHz digital oscilloscope (Tanaka et al., 2010). The timing control precision in the whole system was 8.4 ps (in rms), which was estimated by time-resolved diffraction measurement using a fast lattice response of semiconductor single crystal. Since the distribution of the crossing points is narrower than the pulse width of the SR, the pump-probe system has a time resolution of < 40 ps.

Fig. 6. High precision RF phase control circuit. The circuit provides a delayed RF signal from the output port with a precision of better than 3 ps (in rms) which is independent of the delay time.

Time Resolved Investigation

**2.4 High-precision X-ray diffractometer** 

oscilloscope.

one sample disk.

the high-precision diffractometer.

of Fast Phase-Change Phenomena in Rewritable Optical Recording Media 265

clearance is too small on the right side due to a relatively shorter working distance of the lens for the pump laser. The reflected light intensity was monitored through a beam splitter with a photodiode having a response time of 10 ns, and the signal was stored in a digital

The intense X-ray beam produced from an undulator is guided to the sample through a double curved mirror system, a pulse selector (rotating shutter), a Si(111) channel-cut monochromator, and a high-precision diffractometer with a phase zone plate (ZP) focusing system (Yasuda et al., 2008), as shown in Fig. 8. The ZP has the diameter of 100 μm, the innermost zone radius of 5.0 μm, the outermost zone width of 250 nm, and the tantalum thickness of 2.5 μm. The photon energy, *E*, and energy resolution, *Δ E*/*E*, tuned by the channel-cut monochromator, were 15 keV, and ~10-4, respectively. The X-ray beam was focused by the ZP, and the 1st order focused beam is only selected with the order sorting aperture placed in front of the sample position. The beam size at the sample position was ~3 μm, evaluated by a knife edge scan method using Au meshes. The focused beam intensity for one pulse at the sample position was ~104 photons/pulse, which is ~10-11 of the number of photons for a second in the upstream of the pulse selector. As described in Table 1, the total number of photons incident on the sample are then estimated to be ~1010 photons for

Fig. 8. Photograph of the X-ray pulse selector, the Si(111) channel-cut monochromator and

A curved imaging plate (IP) was employed to obtain an X-ray diffraction profile. The use of an IP is advantageous in such longer exposure time as ~30 min in the present experiment. To obtain one-dimensional diffraction profile, the signal intensity was integrated using the

full area of IP to compensate the smaller number of incident photons on the sample.

#### **2.2 Rotating stage for sample holder**

Repetitive measurement is necessary when the diffraction intensity for one shot was too low to obtain the profiles with high time resolution. Therefore the sample was rotated to give a virgin sample for every measurement, as shown in Fig. 2. To perform 1 kHz-repetitive measurement, the rotation speed of the disk and the laser beam size on the sample are adjusted to satisfy the requirement of v (m/s) > f (Hz)・s (m), where the v, f, and s are the moving speed of the disk, the repetition rate, and a beam diameter of the laser or X-rays, respectively. In addition, an X-ray microbeam technique with a phase zone plate was applied to achieve a few micron X-ray beam size, since the smaller beam gives a larger number of data for one disk. The laser beam diameter on the sample was adjusted to be about 30 μm in the moving direction with a 40 mm-focal lens. The parameters in the experiment are summarized in Table 1, together with the parameters for 5 Hz-measurement without X-ray microbeam (Fukuyama et al., 2008a) for comparison.



Figure 7 shows the photograph of the DVD sample attached on a rotating stage. The measurement with high repetition rate using microbeams enabled us to obtain a diffraction intensity profile within an hour for one disk sample, and to get systematic data.

Fig. 7. Photograph around the sample disk.

#### **2.3 Optical reflectivity measurement system**

On-line monitor system of the time profile of optical reflectivity was installed to check the phase-change condition of the DVD sample. As shown in Fig. 7, a He-Ne laser beam was guided through the transparent substrate onto the backside of the sample, since the clearance is too small on the right side due to a relatively shorter working distance of the lens for the pump laser. The reflected light intensity was monitored through a beam splitter with a photodiode having a response time of 10 ns, and the signal was stored in a digital oscilloscope.
