banner image
No data available.
Please log in to see this content.
You have no subscription access to this content.
No metrics data to plot.
The attempt to load metrics for this article has failed.
The attempt to plot a graph for these metrics has failed.
Crystallization of Ge2Sb2Te5 films by amplified femtosecond optical pulses
Rent this article for


Image of FIG. 1.
FIG. 1.

Optical microscope image of the marks obtained after excitation by single pulses of 60 fs duration. The value of the spatially averaged pulse fluence Favg (mJ/cm2) is shown.

Image of FIG. 2.
FIG. 2.

Optical microscope images of the marks obtained after excitation by a series of 2000 pulses of 60 fs duration. The value of the spatially averaged pulse fluence Favg (mJ/cm2) is shown.

Image of FIG. 3.
FIG. 3.

The change in reflectance of Ge2Sb2Te5 films ΔR/Ram (defined as ΔR = R−Ram where R is the measured reflectance and Ram is the reflectance of the as-deposited amorphous material) after exposure to 2000 pulses of 60 fs pulse width is shown for different values of Favg (the spatially averaged fluence). Spatial profiles of marks recorded for Favg values of 3.50 and 5.83 mJ/cm2 are shown in panels (a) and (b) respectively. Panel (c) shows the measured variation of the maximum reflectance within the area of the mark (squares), the reflectance at the centre of the mark (circles), and relative change of reflectivity measured by the probe beam (triangles) compared with the theoretical reflectance change at the centre of the mark (solid and dashed lines). The dependence of the area of thehigh reflectance region upon Favg is shown in (d), where the black squares are experimental data, the red solid curve is from the geometrical fit and the dashed lines are calculated from the crystallization model.

Image of FIG. 4.
FIG. 4.

The dependence of crystallisation upon the number of 60 fs pulses and Favg . (a) microscope images of marks where the values of Favg have been labelled and the number of pulses varies from 200 at the far left to 600 at the far right with increment of 50. In (b), (c) the dependence of the area of the crystallised region, and the relative change of reflectivity, at the high reflectivity crystalline region, respectively upon the number of pulses is shown for different Favg values. The dashed lines in (b) are the calculated mark areas using the thermal and crystallization models. (d) Calculated optical reflectance curves as a function of pulse number corresponding to the experimental curves in (c) from the crystallization model for two reaction orders, n = 1, 3.

Image of FIG. 5.
FIG. 5.

(a) Microscope image of crystallised mark obtained using 2000 pulses of 60 fs duration with Favg  = 3.50 mJ/cm2. (b) Microscope image after re-amorphization using a single 60 fs pulse with Favg  = 9.33 mJ/cm2.

Image of FIG. 6.
FIG. 6.

Calculated transient temperature distributions using the analytical model derived in this work for applied pump fluence of Favg  = 2.33 mJ/cm2 over 60 fs. Plots (a) and (b) show the temperature distribution through the GST layer as a function of normalised thickness at different times when fully amorphous and fully crystalline respectively. Plots (c) and (d) are the calculated the transient temperatures at different points through the 20 nm thick, fully amorphous and fully crystalline GST layer respectively.

Image of FIG. 7.
FIG. 7.

Calculated average fluence Favg (for a single laser pulse) required to induce melting at the centre of the irradiated region at increasing depths into the GST layer for different crystalline volume fractions. The melting threshold fluence Fmelt is indicated at the average fluence of 4.66 mJ/cm2 observed from measurements and calculations in Figs. 3(c) and 3(d). z = 0 nm corresponds to the surface of the GST layer. The required fluence for melting was calculated using the theoretical heat flow model which does not account for melting and movement of the melting front. This leads to the exaggerated large values of Favg near the capping and under-layers (z = 0 and 20 nm respectively). In practice, the melting front would reach and make contact with these interfaces.

Image of FIG. 8.
FIG. 8.

Reflectance line scans along the semi-major axis of the marks shown in Fig. 4(a) produced using different number of 85 femtosecond pulses at Favg  = 4.20 mJ/cm2 showing the growth of the crystalline material from the inner perimeter and central region of the annular rings with increasing pulse number.

Image of FIG. 9.
FIG. 9.

Measurements of the inner diameters of the annuli in Fig. 4(a) for two fluences as functions of increasing laser pulse number. The straight lines fits reveal an estimate of the elevated temperature growth velocity assuming that the growth occurs over a 2 ns period from the application of a laser pulse, during which the maximum temperature reduces to the glass-transition temperature and the amorphous-to-crystalline interface displacements is at its steady-state value.

Image of FIG. 10.
FIG. 10.

Diagram illustrating the two-layer, semi-infinite model of the experimental stack used for evaluating the temperature distribution T 1 in the GST layer. The laser beam energy absorption, identified by the power density term g, is limited to the thickness of the GST layer d.


Generic image for table
Table I.

List of thermal and kinetic parameters for the Ge2Sb2Te5 and ZnS:SiO2 layers used in the calculations.


Article metrics loading...


Full text loading...

This is a required field
Please enter a valid email address
752b84549af89a08dbdd7fdb8b9568b5 journal.articlezxybnytfddd
Scitation: Crystallization of Ge2Sb2Te5 films by amplified femtosecond optical pulses