Excimer laser ablation of aluminum: influence of spot size on ablation rate

Ablation depth of craters was estimated by measuring the difference in z-position between the surface and the bottom of the crater using an optical microscope. Figure 1(A) shows the ablation depth of craters produced at different laser spot diameters as a function of number of laser pulses. Each data point in figure 1(A) represents the average depth of four craters measured at a fluence of 11 J cm⁻² and a repetition rate of 20 Hz. The variation of ablation rate (calculated from the slopes of figure 1(A)) with laser spot diameter is shown in figure 1(B). Ablation rate was found to decrease from 1.04 µm/pulse at a spot size of 10 µm–0.29 µm/pulse at a spot size of 85 µm. Increasing spot size above 85 µm did not show any further decrease in ablation rate. These results agree with those published by Eyett and Bäuerle [19], which investigated the effect of laser beam spot size of XeCl excimer laser (λ  =  308 nm, pulse width  =  11 ns, repetition rate  <  10 Hz, and fluence up to 3.1 J cm⁻²) on the ablation rate of LiNbO₃. Eyett and Bäuerle [19] reported a decrease in ablation rate by a factor of approximately 3 with increasing spot size from 24 to 80 µm, followed by saturation for craters with diameters from 80 to 175 µm. A ten-fold increase of ablation efficiency was also reported when decreasing the spot diameter from 220 µm to 9 µm in the fluence range of 7.5 J cm⁻²–13.2 J cm⁻² using a KrF nanosecond excimer laser (λ  =  248 nm; pulse width  =  20 ns) during percussion drilling through silicon wafer [15].

Zoom In Zoom Out Reset image size

Experimental and theoretical investigations on the effect of laser beam spot size on the ablation efficiency have attributed the observed decrease in ablation rate with increasing laser beam spot size to the increased attenuation of the incident laser energy for smaller spots by the ablation plume [17, 19]. Beuermann et al [22] assumed a linear dependence of ablation rate on the fluence, which is valid in the saturation region above the threshold fluence. They deduced the following relationship between ablation rate R and laser beam spot radius r [22]

$$\begin{eqnarray}&&R={{R}_{o}}/\left(1+\frac{\rho \sigma {{R}_{o}}}{1+\pi h/2r}\right)\end{eqnarray} \tag{ 1 }$$

where R_o is the ablation rate in the absence of plume attenuation; ρ is the density of the plume; σ is the attenuation coefficient; and h is the distance the ablated particles move during the laser pulse. According to equation (1), a nearly linear relationship between ablation rate R and reciprocal of the spot radius r exists [22] (figure 2). For a very large spot radius, the ablation rate can be expressed by [22]

$$\begin{eqnarray}&&{{R}_{\infty}}={{R}_{o}}/\left(1+\rho \sigma {{R}_{o}}\right).\end{eqnarray} \tag{ 2 }$$

Zoom In Zoom Out Reset image size

The dependence of ablation depth on the number of laser pulses for craters produced at different fluence and at a spot size of 35 µm is shown in figure 3. Ablation rate increased from 0.22 µm/pulse at a fluence of 3.04 J cm⁻²–0.49 µm/pulse at a fluence of 15 J cm⁻². Fluence is one of the most important parameters that controls ablation rate and ablation mechanisms. For many materials ablated using laser systems of different properties, ablation rate was found to increase rapidly with increasing fluence at the sample surface until a certain value of fluence is reached above which ablation rate saturates or decreases [20, 23–25]. The saturation or decrease in ablation rate at high fluence was linked to the evolution of a plasma plume, which increases the effect of plasma shielding and consequently increases the attenuation of the incident laser energy [20, 23–25]. Ablation at a fluence close to the ablation threshold is generally characterized by low ablation rate and low debris deposition around the crater. In the case of femtosecond and short-wavelength nanosecond laser pulses, ablation at low fluence is dominated by non-thermal mechanisms, where melting of the material is not severe. As the fluence increases, thermal ablation mechanisms become more predominant, where melting and ejection of large droplets of molten material are characteristic of the ablation [23, 26].

Zoom In Zoom Out Reset image size

LA-ICP-MS is a quasi-nondestructive technique used for elemental and isotopic analyses of solid samples [27–31]. It is a reliable and sensitive technique for the determination of a wide range of element concentrations (i.e. from ultra trace to percent levels) in solids [29–32]. LA-ICP-MS uses a high power pulsed laser to ablate solid samples located in an air-tight ablation cell. The ablated material is transported from the cell by Ar or He carrier gas through tubing to the mass spectrometer for atomization and ionization by the ion source (i.e. the ICP). The ions are separated and detected inside the mass spectrometer according to their mass to charge ratios. The signal intensity, reported in counts per second (CPS), is proportional to the amount of ablated material that reaches the detector after being atomized and ionized. The concentrations of elements in the unknown solid sample are calculated by comparing the element intensity in the unknowns to that in a standard reference material (i.e. the external calibration standard). In this study, we used LA-ICP-MS as a tool to investigate the effect of laser beam spot size on the amount of ablated material, which is dependent on the crater volume and consequently on the crater depth. This was achieved by using an Agilent 7900 quadrupole ICP-MS for measuring the signal intensity of a major element in the target material (e.g. Al) as a function of laser beam spot diameter.

Figure 4(A) shows the variation of LA-ICP-MS signal intensity of ²⁷Al with laser beam spot diameter for craters produced by 100 laser pulses at a fluence of 11 J cm⁻² and a repetition rate of 20 Hz. Each data point in figure 4(A) represents the average signal intensity measured for four craters of the same spot diameter. The signal intensity was found to increase rapidly with increasing spot diameter up to 85 µm due to the rapid increase in the crater volume and the amount of ablated material. The signal intensity did not show significant variation at spot diameters greater than 85 µm. This indicates that the amount of ablated material and crater volume became less dependent on the spot diameter for spot sizes greater than 85 µm, which is in agreement with measurements made using optical microscopy.

Zoom In Zoom Out Reset image size

A crater produced by an excimer laser has a flat-bottom surface and can be approximated by a cylinder of a volume that equals the product of ablation depth and crater area. The dependence of the crater depth on laser beam spot diameter can be obtained from LA-ICP-MS measurements by plotting the average signal intensity per pulse (CPS/pulse) normalized to crater area as a function of laser spot size (figure 4(B)). The area-normalized signal intensity per pulse, which is proportional to the ablation depth/pulse, decreased with laser spot diameter in a similar manner to that observed in the relationship between ablation rate and laser spot size (figure 1(B)). The deviation between figures 1(B) and 4(B) observed for spot sizes greater than 85 µm can be attributed to the following: (i) LA-ICP-MS is a very sensitive technique (~10⁶ CPS/ppb for the Agilent 7900), which means that small changes in the amount of ablated material result in large differences in signal intensity; (ii) we assumed that the actual crater area is the same as that of the laser beam spot size, however, there are likely deviations from this assumption due to laser-matter interaction and thermal effects. In general, the behavior of ablation depth with spot diameter is very similar when measured by two different techniques (i.e. optical microscopy and LA-ICP-MS).

Figure 5 shows SEM images of craters ablated by 10, 50, and 500 laser pulses using different laser spot diameters (10, 20, 40, 110, and 150 µm) at a fluence of 11 J cm⁻² and a repetition rate of 20 Hz. It is clear from the images that the ablation depth of the craters produced by a fixed number of laser pulses is dependent on the laser beam diameter. This agrees with optical microscopy measurements of the crater ablation depth. It can also be seen from the SEM images that the amount of debris deposited and accumulated around the ablation crater is dependent on the number of laser pulses. Increasing the number of laser pulses leads to an increase in the amount of debris deposition and a build up of a crater rim. A crater rim is formed from re-solidified molten material that cannot escape from the crater. Accumulation and deposition of debris around the craters may be attributed to the inefficient removal of the ablated material and its deposition on the sample surface by the action of gravity due to performing the ablation in ambient atmosphere, where the movement of air molecules is negligible. Conducting ablation in a flowing gas environment or under liquid media reduces or eliminates debris deposition around the crater and leads to a clean crater surface [33–35].

Zoom In Zoom Out Reset image size

Figure 6 shows different types of particles observed in the ablated products. The first type, shown in figure 6(A), was observed to accompany craters with raised rims. This type of particle (>5 µm in size) most likely originated from the collapse of the crater rim, which led to deposition of structures of different shapes and sizes inside and outside the ablation crater. The second type of particle (figure 6(B)), had spherical shapes and originated from ejection of molten material from the crater onto the surrounding sample surface. The presence of spherical particles and the evidence of melting observed at the crater edge (figure 7) indicate thermal ablation [36–38]. Two types of ablation mechanisms (thermal and non-thermal) can occur during ablation. Both mechanisms can exist simultaneously but to varying degrees depending on the type of material being ablated and the laser properties [39–41]. Thermal (photothermal) ablation is the dominant ablation mechanism in cases of nanosecond long-wavelength laser pulses due to absorption of the incident laser energy by electrons and its subsequent transfer to the lattice during the laser pulse. Thermal ablation depends mainly on thermal conductivity of the target material, fluence, and the laser pulse width [9, 40, 41]. Non-thermal (photochemical) ablation is the dominant mechanism of material removal in cases of ablation using deep ultra violet (DUV) or femtosecond laser pulses [9, 40]. In non- thermal ablation, the absorbed photon energy is directly used to break the chemical bonds, which leads to material removal without thermal effects [9, 40, 41]. Even with femtosecond or nanosecond DUV laser pulses, thermal ablation becomes significant if optimal laser conditions are not applied [9, 23, 26, 40–42].The third type of particle was much smaller in size than the other two types and was characterized by agglomeration into chain-like structures (figure 6(C)). Condensation from the vapor phase is the most probable origin of formation of this type of particle [36, 38, 43, 44].

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size