Colloidal solutions of nanoparticles obtained by laser ablation are the basic form of nanomaterials, which are used in biomedicine, nanophotonics, materials science and others . The high intensity of periodic pulses of the laser makes it possible to obtain ecologically and efficiently chemically pure colloidal solutions of various types of nanoparticle materials .
In contrast to femtosecond and picosecond laser irradiation, for which phase explosion mechanism predominates [4-6], several mechanims are well known for “dry” conditions in the nanosecond generation mode [5,7-12]. Therefore, there is a need for a proper selection of the generation parameters. As the intensity of the laser radiation increases, the following proceeds occur: 1) surface vaporization of molten materials along their “melt-vapor” binodes, ending up by short condensation of corresponding low-density atomic or small-cluster vapors [5,13]; 2) homogeneous boiling in the proximity of their “melt-vapor” spinodes and almost coinciding with high-density facilitated optical breakdown above the ablated surface, the related onset of sub-critical plasma and ambient shock-wave emission ; 3) the following sub-critical plasma regulates laser energy coupling
(1) surface pressurization
(2) and ablation rate
(3) on the surface described by the well-established scaling relationships as functions of material (atomic mass A, average ion charge Z, ) and laser (intensity I [W/cm ], wavelength [cm] and pulsewidth [s]) parameters  and result in dissociation/ionization of ablation products in the hot plasma core above the surface prior their recombination/condensation [14-15]; 4) finally, deep material melting and superheating by transient bremsstrahlung and recombination plasma emission and its mechanical unloading during plasma adiabatic expansion [17-19] results in intense expulsion of micro-droplets [11-12,20]. Meanwhile, possible appearances of these well-known ns-laser ablation mechanims under wet ablation conditions were not firmly observed and justified yet [1,2].
This work describes on nanosecond-laser plasma-mediated ablation of silicon wafers in water, revealing its underlying fundamental mechanisms by analyzing informative laser intensity dependences for ablative mass yield, SEM analysis and extinction coefficients of generated colloids.
2. Experimental details
For ablation was done scanning of a 0.5-mm thick, 20x20 mm wide commercial monocrystalline Si under a 2-mm deionized water layer for 10 minutes by using a laser marker HTFMARK (Bulat), comprised by a Yb -fiber laser (wavelength – 1070 nm, half-maximum pulsewidth – 120 ns, pulse energy – 1 mJ, repetition rate f = 20-80 kHz), a galvanoscanner (f 160 mm), a motorized translation from PC (Fig.1a). Laser pulses with different pulse energies E=0.4-1 mJ, coming at f = 20 kHz, were focused into a 22-μm wide (1/e-diameter σ ) spot on a wet sample surface (Fig.1b, the peak laser fluence F 265 J/cm an the peak laser intensity I 2.2 GW/cm at E = 1mJ) and scanned across 4x4 mm large area with 10-lines/mm filling at the scan velocity V = 80 mm/s.
3. Experimental results
The single-shot “wet” ablation spots produced at variable peak laser intensities (Fig.1b) exhibit a linear dependence in the D -lnI coordinates with the squared 1/e-diameter of the laser focal spot as the slope σ = (22 1) μm and the single-shot ablation threshold I = 0.79 0.09 GW/cm as the horizontal offset. This high ablation threshold indicates the low energy coupling at the 1-μm laser wavelength to the Si sample near its indirect absoprtion edge until the material melting, yielding the strongly absorbing molten phase. The corresponding two-photon absorption (the absorption coefficient 2 cm/GW in low-intensity experiments or 50 cm/GW in high-intensity experiments), which is necessary to initiate free-carrier absorption and avalanche ionization , becomes considerable for the 120-ns long laser pulse only at high laser intensities, comparable to I . At higher intensities I I 1.5 GW/cm (the shadowed region in Fig.1b), the single-shot ablation exhibits the strong dispersion and saturation of this D -lnI curve, apparently, related to lateral plasma screening.
The obtained dependence of the sample mass loss at different intensities I I (Fig.2, left axis) demonstrates in the range I = 0.9-1.5 GW/cm a sublinear dependence M I , which resembles with its slope M 0.7 the sub-critical plasma-mediated ablation rate dependence in Eq.(3). In contrast, a threshold-like, non-linear mass-loss increase ( M I ) almost by one order of magnitude occurs at higher intensities I 1.5 GW/cm (the shadowed region in Fig.2), i.e., above the second multi-shot ablation threshold I . In accordance with the previous “dry” ns-laser ablation studies, temporaly delayed expulsion of micron-sized melt droplets is expected in this second intensity range as a result of plasma-driven “bulk” phase explosion, abruptly increasing the ablation rate [12,17-20].
Finally, the transmitance of the Si-NP hydrosols, rather monotonously increasing over the spectral range of 550-1150 nm (Fig.1d), was sampled in two spectral ranges – about 700 and 1000 nm – and converted to their corresponding extinction coefficients [cm ] = ln T (700,1000)/lnT (700,1000) for the 1-cm long quartz cuvette (Fig.3). This quantity demonstrates for both these spectral ranges a rapid – third-power – increase above the ablation threshold I , eventually tending to saturation for I 1.5 GW/cm (rather similar, threshold-like trend with non-linear arise and subsequent flattening was observed for colloidal extinction). This observation is also consistent with expulsion of micro-droplets at higher laser intensities, with such micro-droplets negligibly contributing to the colloidal extinction.
This apparently indicates that almost total redeposition can occur during the near-surface vapor buble collapse in the confined ablation regime, making nanoparticle extraction from the bubble to the ambient liquid the second important task after the ablative removal itself. Such extraction can be potentially enhanced by 1) overlapping the consequent bubbles on the surface prior collapse of the previous one and flushing (transferring) its content to the next, emerging one, being managable in the scanning regime for energy-dependent (sub)ms bubble (oscillation) lifetimes T and (sub)mm-wide bubble dimensions D  by changing laser repetition rate and scanning velocity to satisfy the condition D f/V 1 (apparently, this is the case in this study); 2) multi-shot repetitive ablation in the spot-by-spot ablation regime, removing by the next laser shot in each spot the material redeposited upon the previous bubble collapse; 3) multi-shot ablation at T f 1, continuously accumulating the ablated matter inside the persistent bubble.
In this study, two modes of nanosecond-laser ablation of crystalline silicon wafer in water in plasma-mediated regime were revealed by investigating mass removal and generation of corresponding colloidal solutions versus increasing laser intensity. Plasma-mediated ablation is supposed to be initiated by laser-induced “surface” phase explosion via optical breakdown of dense, hot ablative plume and results just above the ablation thresholds in characteristic sub-linear mass removal yield and third-power raise extinction of colloidal solution versus laser intensity, indicating their regulation by near-surface sub-critical plasma via surface screening and dissociation of ablation products in the hot plasma core. At higher intensities a threshold-like onset of plasma-induced bulk phase explosion is observed as a drastic raise of mass removal in the form of melt micro-droplets and the corresponding saturation of extinction dependence on laser intensity. Our findings suggest the significant role of near-surface nanosecond-laser ablative plasma in regulating primary mass and energy input, as well as driving plasma pressure inside the succeeding vapor bubble in water, which manages the accompanying nanoparticle formation/transformation dynamics.