Rarefaction after fast laser heating of a thin metal film on a glass mount

Understanding the physics of laser-matter interactions in ultrashort pulses is important for many well-acknowledged applications from material modifications to biology. We numerically and experimentally consider the effect of sub-picosecond Ti:sapp laser actions on 60-100 nm gold films mounted onto a fused silica substrate. The pulse energy is sufficient to ablate the films. For the first time, we show that there are different regimes of ablation, and the formation of the 3D structures depends on the value of the absorbed fluence F-abs and the adhesion strength p(adh) between the film and the substrate. Namely, a delamination threshold F-delam and an ablation threshold F-abl (F-delam < F-abl) exist if adhesion is weak. Above the lower threshold F-delam, the whole film delaminates from the substrate. Above the higher threshold F-abl, the thin film ruptures near its middle plane. The external half of the film flies away, while the internal half remains on the glass substrate. There are two thresholds F-delam and F-abl for the Au/glass and Ag/glass targets, because pure gold and silver are weakly coupled to the glass. The lower threshold Fdelam disappears in the case of a strong adhesion stress p(adh). Consequently, delamination as a whole becomes impossible. Although the rupture of a film remains because the ablation threshold F-abl is independent of on adhesion. Adhesion is high when an intermediate thin layer of chromium is placed between the gold and the glass. The film velocity after it separates from the substrate is low for the range of fluences F-delam < F < F-abl because the acoustic impedance Z(glass) of glass is small relative to the impedance Z(film) of a gold film. Therefore, during an evolution, the absolute values (positive or negative) of the contact pressures are few times smaller than the absolute pressures in a gold film. However, above the ablation threshold F-abl < F, the velocity of the external part of a film becomes few times larger because the velocity is independent of the acoustic impedance of a substrate. These circumstances explain why 3D structures such as nanojet above the microbump appears in the case of a weak adhesion. The velocity should be not too large to allow to the surface tension and crystallization to stop an inflation and breakaway of a microbump.