Ultrafast x-ray diffraction study of melt-front dynamics in polycrystalline thin films

Abstract

Melting is a fundamental process of matter that is still not fully understood at the microscopic level. Here, we use time-resolved x-ray diffraction to examine the ultrafast melting of polycrystalline gold thin films using an optical laser pump followed by a delayed hard x-ray probe pulse. We observe the formation of an intermediate new diffraction peak, which we attribute to material trapped between the solid and melted states, that forms 50 ps after laser excitation and persists beyond 500 ps. The peak width grows rapidly for 50 ps and then narrows distinctly at longer time scales. We attribute this to a melting band originating from the grain boundaries and propagating into the grains. Our observation of this intermediate state has implications for the use of ultrafast lasers for ablation during pulsed laser deposition.

Fig. 1. Experimental configuration for TR XRD from polycrystalline gold thin films.

(A) Experimental setup showing the sample, the Rayonix detector MX225-HS, and the photodiode (I₀) to measure the transmitted beam for normalization. The sample was mounted perpendicular to the 9.7-keV x-ray beam, and a 400-nm, 100-fs optical laser was used to excite the sample in almost colinear geometry. (B) 2D diffraction patterns of the 300-nm thin film collected 50 ps before and 100, 220, and 390 ps after laser excitation at an incident laser fluence of 254 mJ/cm². (C) Cross-sectional view of the gold thin film and substrate window array arrangement.

Fig. 2. Dynamics of azimuthally integrated XRD profiles of different peaks.

(A) (111) (B) (200), and (C) (311) diffraction peaks for the 300-nm-thick thin film measured at various delay times after an incident laser fluence excitation of 254 mJ/cm². All the diffraction positions have both the crystal powder and intermediate peaks. a.u., arbitrary units.

Fig. 3. Dynamics of the (111) crystal diffraction peak measured from 300-nm gold thin films.

(A) Fits to the diffraction intensity profiles versus momentum transfer, Q, measured 50 ps before and 268 ps after excitation at an incident laser fluence of 254 mJ/cm². (B) Peak position and (C) integrated intensity as a function of time delay. For the 254 mJ/cm² data, the peak position was fitted with a sum of exponential decay (time constant, τ₁ = 2800 ± 50 ps) and an exponentially damped cosine function (with damping time constant, τ₂ = 90 ± 10 ps, and period, T = 130 ± 10 ps). Similarly, an integrated intensity of 254 mJ/cm² was fitted with τ₁ = 2800 ± 100 ps, τ₂ = 250 ± 45 ps, and T = 130 ± 10 ps.

Fig. 4. Dynamics of the intermediate diffraction peak of the (111) profile.

(A) Integrated intensity, (B) integrated intensity of the intermediate peak at different incident fluences, (C) position, and (D) width as a function of pump-probe delay time for 300-nm films. The data measured at different incident laser fluences are indicated with different colors. The peak width was fitted with an exponential decay convoluted with a Gaussian function. For the data measured at 254 and 127 mJ/cm², the time constants were 1320 ± 50 ps and 3900 ± 600 ps, respectively. The peak position measured at 254 mJ/cm² was fitted to a sum of two exponential functions with time constants of 50 ± 10 ps and 330 ± 20 ps convoluted with a Gaussian function. The dashed blue line shown in (C) is the expected position of the gold (111) peak at the melting point due to thermal expansion, assuming ambient pressure.

Fig. 5. Thickness dependence of the melting time of gold thin films.

(A) Diffraction intensity profiles measured from a 50-nm-thick gold film at an incident laser fluence of 50 mJ/cm² at different delay times after laser excitation. (B) Time dependence of the integrated diffraction intensity of 50-nm thin films at different incident laser fluences. (C) Diffraction profiles of a 100-nm-thick gold film measured at an incident laser fluence of 66 mJ/cm². (D) Integrated intensity of the (111) peak as a function of pump-probe delay time for different fluences. (E) Time dependence of the crystal peak component for the 300-nm gold thin films for different incident laser fluences. (F) Residual crystal fraction versus fluence for 100- and 300-nm-thick films. The fraction of the crystal peak intensity remaining 500 ps after laser excitation is plotted versus incident laser fluence and fitted with a sigmoid function.

Fig. 6. GB melting mechanism in a polycrystalline gold thin film.

(A) Sketch of GB locations in a polycrystalline gold thin film and a zoomed view of how the melt front would propagate away from the GB following optical laser excitation. (B) Simulation of the spatial temperature distribution using the heat diffusion equation from a spike of melt created at the GB, at x = 0. The heat moves rapidly into the grain with a melt-front velocity determined by the heat flux. The shaded offset region is due to the uptake of latent heat, resulting in a block of melting material sandwiched between two melt fronts moving at different velocities.