Contents
4.1 Introduction . . . 120 4.2 Pump-probe experimental test-bench. . . 121 4.3 Sample positioning. Focusing pump and probe beams. . . 123 4.3.1 Spatial overlapping of the laser spots on the sample surface. . . 123 4.3.2 Generation of the second harmonic of the laser for superposition two
laser beams temporarily. . . 123 4.4 Reflection, transmission and absorption as functions of time in the-
ory and experiment. . . 129 4.4.1 Energy balance. Distinction between damage and ablation thresholds. . 132 4.4.2 Moment of the beginning and end of absorption. Absorption duration. . 133 4.4.3 Comparison of the theoretical and experimental time scales for energy
deposition . . . 134 4.4.4 Efficiency of the 500 fs laser, optimal F . . . 135 4.5 Conclusion . . . 135
than knowledge. Knowledge is limited. Imagination encircles the world." Albert Einstein
4.1
Introduction
In the previous chapter we studied the reflection, transmission and absorption dependence on energy and time. We have demonstrated that the dynamics of absorption and plasma shielding processes are energy (fluence) dependent. In particular, we have defined an optimal fluence window yielding high efficiency of energy deposition. It is now interesting to estimate the quality of micromachining (especially in this fluence range) to validate this process of surface direct laser-writing.
In the following we discuss the morphology of ablation craters and also the interest to use, in terms of ablation selectivity, a 500 fs laser for dielectric material treatment.
The last part of this chapter is devoted to the study of a medical application of subpicosecond lasers. We develop a parametric study to optimize the laser parameters for reproductible corneal grafting. This study aims to find and to fix the ensemble of laser parameters to benefit from the special feature of a femtosecond laser such as the precision of the energy deposition and the minimal energy dose for minimizing postoperative complications and facilitating tissue healing.
4.2
Application of femtosecond lasers for drilling
Concerning the applicative aspect of the present work, let us consider the selectivity of laser ablation in the wide range of fluences explored. We present the ablation characteristics (ablated diameter, volume and depth) in fig 5.1 and the topology of the ablation craters in fig 5.2 and in fig. 5.3. All data are normalized with respect to the damage threshold (4.4 J/cm2).
From fig. 5.1a, we see that the ablated diameter never achieves the total diameter size defined as d = πω0. From ∼ 2 F/Fth there is a linear increase of the ablation diameter up to ∼ 6
F/Fth. Moreover, the crater diameter does not depend on intensity but only on normalized
fluence F/Fth as it is shown in ref. [114]. Hence using a small increment in fluence opens the
possibility to control with a one micron precision the transversal size of the ablated holes by thoroughly choosing applied fluence. Note however that subpicosecond laser pulses having more pronounced thermo-mechanical effects compared to ultrashort laser pulses (electronic effects) could have a depressed rim surrounding ablation crater [28]. The quality of ablation at low fluence is characterized by a higher level of debris and loss of regularity, but with fluence increase the crater walls and bottom becomes free of debris (see fig 5.2 and in fig. 5.3) that are probably ejected far from the target. The fluence range where the ablation is efficient is determined as ∼ 90 % of Fmax,eff1 (see fig. 3.9 in section 3.2.2.2). According to that criterion, the ablation is
1
efficient from 3F/Fth to 12.5F/Fth (in terms of laser fluence from 13 to 55 J/cm2).
Figure 4.1: Ablation characteristics as a function of normalized laser fluence. a) Ablated crater diameter vsF/Fth. b) Ablated crater volume vsF/Fth. c) Ablated crater depth vsF/Fth. Error
Figure 4.2: Three dimensional AFM snapshots (left) and one dimensional profiles (right) for three laser fluences. The laser fluences 1.7F/Fth and 2.8 F/Fth are below and 5.2 F/Fth is in
the range of optimal fluences for micromachining as it is defined from fig. 3.9 in section 3.2.2.2. The total beam diameterd = πω0 = 19.8 µm.
Figure 4.3: Three dimensional AFM snapshots (left) and 1D profiles (right) for two fluences. The laser fluences 8.8 F/Fth is in the range of optimal fluences for micromachining and 16.2
F/Fth is above this range. The total beam diameter d = πω0 = 19.8µm.
The fluence range (from 1.7 F/Fth to 16.2 F/Fth) analyzed here starts from values below
the optimal fluence and finishes when the ablation becomes again non-efficient. The shape and morphology of ablated crater is irregular for 1.7F/Fth(see fig. 5.2) close to ablation threshold (1.3
F/Fth). The crater shape is Gaussian-like with high roughness of the crater walls and bottom.
The maximal depth achieved for this fluence is in order of 65 nm and the crater diameter is ∼ 4.5 µm. When the incident fluence increases, the crater depth and diameters increases, but for low fluence (2.8 F/Fth, see fig. 5.2) high irregularity is still present. The crater shape is also
Gaussian-like similar to the previous case.
We now examine the laser fluence that is in the range of optimal micromachining regime. In the range of optimal laser fluences all ablation characteristics increase with the normalized fluence. Ablated depth increases swiftly up to 3 F/Fth, where the control in depth increment
becomes comparable with ∼ 20 nm. The maximal depth achieved at 12.5 F/Fth equals to ∼
300 nm. By varying incident laser fluence and considering the focusing used here (NA = 0.09), one can reach different depths from ∼ 50 to 300 nm (up to 0.003zr) and diameters from ∼ 4 to
13 µm (up to 2.06ω0) using a subpicosecond laser. Indeed, femtosecond laser possesses highly
deterministic and precise nature. Note that it could be even possible to have a better resolution (ablation selectivity) with a more stable laser system (in our case ∼ 5% rms pulse to pulse energy fluctuations). For the laser fluence in the range of optimal ablation efficiency (5.2 F/Fth, see
fig. 5.2), the crater depth increases further, but the difference with two lower fluences consists in smoother bottom surface and crater walls. There is no influence of the plasma shielding and the laser pulse absorption is more efficient in the center of the beam where the local intensity is higher yielding Gaussian-like crater shape. With further fluence increase, the depth saturation causes a transition from Gaussian-like to "Top-hat" shape of the resulting ablation crater (8.8F/Fth, see
fig. 5.3). The progressive change of the crater shape in the range of optimal ablation efficiency hence allows to access to particular shape, i.e. Gaussian-like or "Top-hat" only be adjusting the incident laser fluence. For high fluence cases (8.8 F/Fth and 16.2 F/Fth), the crater depth
saturates to 250 nm, but the crater diameter still increases. We therefore conclude that the axial resolution (crater depth) depends more on local fluence2 and the crater diameter depends on total fluence. The quality of the crater walls is high, they are steep and smooth. For 8.8F/Fth,
a small bump appears in the crater bottom. It is probably related to the plasma mirror effect that is reached in the region of the highest local intensity. This bump becomes more pronounced at fluence 16.2 F/Fth and causes the degradation of the quality of laser ablation. It is thus
important to work in the optimal fluence window range, since here the maximal absorption is attained and the energy deposition is not significantly disturbed by the plasma shielding effect. A subpicosecond laser therefore has one important advantage over ultra short laser pulses, the working window is wider than for ultra short pulses [114].