Short-pulse UV laser ablation of solid and liquid metals: indium

Laser-ablation experiments on solid (T = 300 K) and liquid (T = 600 K) indium are reported. The ablation was performed under high vacuum conditions with UV laser pulses at 248 nm and pulse durations of 15 ns and 0.5 ps. The ablated neutral indium atoms were resonantly ionized with a second laser pulse 28 mm above the sample surface and detected in a time-of-flight mass spectrometer. The ablation threshold fluence for solid indium decreases by a factor of 40 from 100 mJ/cm2 to 2.5 mJ/cm2 when a 0.5 ps pulse is used instead of a 15 ns laser pulse. Measurements on liquid indium show a different behavior. With 15 ns laser pulses the threshold fluence is lowered by a factor of ∼ 3 from 100 mJ/cm2 for solid indium to 30 mJ/cm2 for liquid indium. In contrast, measurements with 0.5 ps laser pulses do not show any change in the ablation threshold and are independent of the phase of the metal at 2.5 mJ/cm2. This behavior could be explained by thermal diffusion and heat conduction during the laser pulse and demonstrates in an independent way the energy lost into the material when long laser pulses are applied. Time-of-flight measurements to investigate the underlying ablation mechanism show thermal behavior of the ablated indium atoms for both ps and ns ablation and can be fitted to Maxwell–Boltzmann distributions. PACS: 78.66.Bz; 42.62.-b; 81.15.Fg The interaction of laser light with materials and the processing of surface modifications have been investigated for many years. Nevertheless the study of desorption and ablation processes is still a growing field in basic science, engineering, and micro-mechanics [1–6]. The study of the basic desorption and ablation mechanisms is a great challenge, since the underlying processes are still not completely understood. On the other hand, laser material treatment is applied in very different fields, especially in micro-machining and micro-fabrication. Obviously, laser ablation depends on a large number of parameters such as the material properties, laser parameters, and chemical environment. Therefore, the investigation of the suitable regimes for laser treatment of different materials is of high priority. Current laser ablation experiments concern semiconductors [7], materials such as crystals and glasses [8–10] that are transparent in the visible, polymers [11, 12], and metals [13, 14]. The precise ablation of metals presents some difficulties, since the high thermal conductivity of metals results in a large energy loss by heat diffusion into the bulk during the laser pulse. This dissipating thermal energy leads to an increasing temperature of the material outside the laser spot, not resulting in ablation but modification of the crystal structure. For this reason, it is important to reduce the energy loss and heat transfer into the material, to prevent undesired modifications, and to search for new nonthermal, i.e. electronic, ablation mechanisms. One possibility for reducing the energy loss into the material is to shorten the laser pulse duration to the femtosecond region. For fs laser pulses the heat diffusion into the material is negligible and, as a consequence, the energy loss during the laser pulse is minimized [15, 17]. The thermal diffusion length Lth is given by Lth ∝ (κτ p), wereκ is the thermal diffusivity and τ p is the pulse duration. The thermal diffusivity is related to the thermal conductivity k byκ = k/%c, were% is the density and c is the heat capacity of the metal [16]. For 15-ns pulses the thermal diffusion length for metals is of the order of 1 μm and decreases to about 20 nm for 0.5-ps pulses [17]. Consequently, rapid energy accumulation occurs in the optical absorption volume for fs pulses and hence efficient ablation takes place if the accumulated energy reaches a critical value, of the order of the enthalpy of evaporation. In other words, compared to the case with ns laser pulses, the ablation threshold is drastically lowered for short laser pulses. To measure the influence of the thermal heat diffusion on the ablation process we have investigated the ablation threshold fluence directly by changing the laser pulse duration from 15 ns to 0.5 ps. A second and independent way to show the influence of energy loss by heat diffusion into the material is, according to the relation Lth ∝ (κτ p), to vary the thermal properties, i.e. the thermal diffusivity κ or the thermal conductivity k at a constant laser pulse duration. Since the thermal conductivity for indium at the melting point changes in a step-like function by a factor of 2 [18–20], this must influence the thermal diffusion length and hence the energy loss into the material during the laser pulse.