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

264 Materials Science and Technology

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.

Laser beam size (μm)

intensity profile within an hour for one disk sample, and to get systematic data.

1 × 103 8 × 10-2 30 3 1.8 × 106 5 5 × 10-3 300 50 3 × 104 Table 1. Optical parameters of the time-resolved X-ray diffraction measurement system.

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

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

X-ray beam size (μm)

Number of repetition for one disk

Disk rotation speed (m/s)

Fig. 7. Photograph around the sample disk.

**2.3 Optical reflectivity measurement system** 

**2.2 Rotating stage for sample holder** 

Repetition rate (Hz)

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 one sample disk.

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

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.

Time Resolved Investigation

properties.

and (b) AIST.

of Fast Phase-Change Phenomena in Rewritable Optical Recording Media 267

accordance, which prove the close relation between the X-ray diffraction intensity and the optical reflectivity of the phase-change materials, *i.e.,* the structure and the electronic

Fig. 10. Time-evolution of the optical reflectivity and X-ray diffraction intensity of (a) GST

From the intensity profile measured by the APD/MCS method, the delay times, *τ*, were determined for the IP/pump-probe method. Figure 11(a) shows the diffraction patterns obtained from the IP/pump-probe method for a 40 ps snapshot. Since the intensity of each diffraction peak has a uniform time-dependent increase, there is no crystal-crystal phase transition in GST and AIST during the crystal growth. We estimated the grain sizes from the line width of Bragg reflection, as shown in Fig. 11(b). In GST, the grain size is nearly constant (=70 nm), while the grain size significantly increases up to 1 μs in AIST. It is remarkable that the volume fraction of the crystal phase is almost saturated at 300 ns (see Fig. 10). Thus these observations suggest the coalescence of the crystal domains after 300 ns.

Fig. 11. (a) Snapshots of X-ray diffraction patterns and (b) the changes in peak width.

From the above experimental findings, we propose schematic models for the crystallization processes of GST and AIST as shown in Fig. 12. In the case of GST, nucleation takes place in the whole area in the amorphous phase after laser irradiation (A), and the number of newly formed crystallites of 70 nm diameter increases during the cooling process until 300 ns (B-

#### **2.5 Performance of the X-ray pinpoint structural measurement system for investigation of optical recording process**

Here we briefly summarize the X-ray pinpoint structural measurement system for investigation of optical recording process (see Fig. 2). The time-resolution and time precision are ~40 ps and ±8.4 ps, respectively. The spatial resolution is determined by the area of the laser beam of 30 μm, which is larger than the X-ray beam size of 3 μm. The maximum repetition rate is 1 kHz, in which the fresh sample area on the rotating disk appears at the irradiating position for every measurement. All parameters shown here are described in full width at half maximum.
