Femtosecond and picosecond laser microablation: ablation efficiency and laser microplasma expansion

Share Embed


Descripción

Appl. Phys. A 69 [Suppl.], S381–S383 (1999) / Digital Object Identifier (DOI) 10.1007/s003399900230

Applied Physics A Materials Science & Processing  Springer-Verlag 1999

Femtosecond and picosecond laser microablation: ablation efficiency and laser microplasma expansion B. Sall´e1 , O. Gobert2 , P. Meynadier2 , M. Perdrix2 , G. Petite3 , A. Semerok1,∗ 1 CEA

Saclay, DPE/SPCP/LSLA, Bât.391, 91191 Gif sur Yvette cedex, France (Fax: +33-1/6908-7738, E-mail: [email protected]) Saclay, DSM/DRECAM/SPAM, Bât.522, 91191 Gif sur Yvette cedex, France Saclay, DSM/DRECAM/SRSIM, Bât.462, 91191 Gif sur Yvette cedex, France

2 CEA 3 CEA

Received: 21 July 1999/Accepted: 31 August 1999/Published online: 22 December 1999

Abstract. Laser ablation efficiency and plasma plume expansion were studied using the interaction of Ti-Al2 O3 laser pulses (wavelength 800 nm; energy 20 µJ; mode TEM00 ; waist diameter 11 µm; pulse durations 70 fs, 150 fs, 0.4 ps, 0.8 ps, 2 ps, and 10 ps) with copper in air. A moderate laser pulse energy of 20 µJ was used to eliminate the sharply focused femtosecond laser beam disturbance caused by its nonlinear interaction with air. The craters formed at the surfaces were measured with 0.1 µm longitudinal and 0.5 µm transverse resolution. Laser plasma expansion was measured by an ICCD camera with 3 µm spatial and 1 ns temporal resolution. These measurements were performed in a time delay range of 0–50 ns. The laser pulse duration range used in our study was of particular interest as it corresponded to the characteristic time for electron–phonon interactions in solids (of the order of one picosecond). Thus we could study the different regimes of laser ablation without (fs pulses) and with (ps pulses) laser beam/plasma plume interaction. Laser ablation efficiencies, crater profiles, plasma plume shapes at different time delays, and rates of plasma expansion in both longitudinal and transverse directions to the laser beam were obtained for all the laser pulse durations mentioned above. The experimental results of our investigation on laser ablation with short laser pulses were analysed from the point of view of different theoretical models of laser beam interaction with plasma and metallic surfaces. PACS: 79.20.Ds

A powerful laser beam focused on a solid target results in a crater formation on the target surface and the creation of a plasma composed of electrons, excited atoms, and ions. The plasma obtained can be studied by Optical Emission Spectroscopy to determine the target composition. To make the ∗ Corresponding

author.

COLA’99 – 5th International Conference on Laser Ablation, July 19–23, 1999 in Göttingen, Germany

analysis more accurate, a knowledge of laser ablation properties and mechanisms is important. The interaction process is a complex phenomenon depending on the laser beam parameters (pulse duration, energy, wavelength, angular divergence, spot size), the physical properties of the solid target, the surrounding environment’s composition, and the pressure. The laser pulse’s interaction with the near-surface plasma can affect the laser beam distribution on the target surface. Thus the description of craters (diameter, depth, volume, and shape) and plasma plume expansion is seen as an interesting method of understanding the principal physical mechanisms of the laser ablation process. The aim of this work was to investigate the ablation efficiency and microplasma expansion with pulses in a range from 70 fs to 10 ps. Thus the different regimes of laser ablation could be studied. They were distinguished by the characteristic time of electron- phonon interaction in solids (of the order of one picosecond) and were analogous to ablation without (fs pulses) and with (ps pulses) laser-plasma interaction [1–4]. 1 Experiments The experiments were carried out in air with a pulsed laser with 800 nm central wavelength. The laser operation principle was based on chirped pulse amplification [5] in titaniumsapphire crystals. Adjustment of the distance between the two gratings of the compressor allowed us to choose the pulse duration in the range 70 fs to 10 ps. Pulse durations of 70 fs, 150 fs, 0.4 ps, 0.8 ps, 2 ps, and 10 ps were applied. Each value was defined with the second-order autocorrelator [6]. The contrast of the pulse was measured with a high dynamic cross-correlator [7]. The rate of amplified spontaneous emission was estimated to be less than 10%. Spatial filtering was used to obtain a nearly TEM00 laser beam. Beam analyser was used to determine the energy distribution profile of the laser beam. The waist diameter (FWHM) was found to be 11 µm at 150 mm lens focus. The copper sample was located in the focus plane of the lens. The adjustment of the sample position relative to the lens was made by measuring the crater size, with the assumption that the

S382

the pulse number increases, the crater depth grows linearly while the crater diameter remains constant. Thus, with increasing pulse number the crater volume demonstrates a behavior similar to that of the crater depth. Craters formed by pulses of different durations are of identical typical profile. No ablated matter (“corona”) is detected on the target surface around the crater. Figure 2 presents the crater depth vs. the laser pulse duration. No significant change in crater depth is observed in the range 70–800 fs of laser pulse duration. For higher pulse durations, the crater depth decreases with increasing laser pulse duration. Similar behavior is observed for the crater volume vs. laser pulse duration. For all the pulse durations used in our experiments, the plasma created after the second laser pulse underwent a similar typical evolution. The longitudinal and transverse plasma dimensions could be derived from the plasma images. Figure 3 presents the plasma’s longitudinal dimension. The transverse dimension experienced an identical temporal evolution. During the first nanoseconds, the ablated matter was observed to escape from the surface with a fast expansion in the direction normal to the target. Rates of plasma plume expansion for different laser pulse durations could be determined from the plasma images. The initial rates of plasma plume expansion were not found to depend significantly on the laser pulse duration. The longitudinal rate of expansion was determined to be about 4.6 × 105 cm/s and the transverse rate was found to be about 3 × 105 cm/s. These values were close to the velocity of sound in the copper target. After 25 ns there was no further evolution of the plasma volume.

Depth / pulse (µm)

most appropriate sample position should provide the smallest crater diameter. To avoid nonlinear propagation effects in air (near the focus zone with very short pulses), we chose an energy per pulse of 20 µJ, corresponding to a fluence of 21 J/cm2 . The normal repetition rate of the laser was of 20 Hz. In our experiments, an electromechanical shutter allowed us to choose the desired number of pulses for crater formation. The craters formed on the copper surface after 2, 5, 10, and 20 laser shots were studied with an optical microscope profilometer (MicroXam Phase Shift Technology) of 0.5 µm lateral resolution and 0.1 µm longitudinal resolution. Crater volumes were measured with reference to the sample surface. The laser plasma expansion in the time delay range of 0–50 ns was obtained with an intensified gated CCD camera (Hamamatsu C4346-01, 756 × 581 pixels, spectral range ≈ 200–800 nm) with a gate time of 3 ns. The gating of the CCD camera was performed by a digital gating system (Stanford Research Systems DG 535) synchronized by the laser trigger output. The time delay for image registration was measured relative to the onset of the laser pulse and was applied with a one nanosecond step. In this case, the experimental temporal resolution of plasma plume expansion was 1 ns. A microscope objective of 40× magnification was used for imaging of the laser plasma with 3 µm lateral resolution. One laser pulse was used for pre-ablation and the plasma image was recorded after the second laser pulse. The plasma plume luminosity was found to depend on the delay time. To avoid CCD camera saturation, we changed the intensifier gain before each plasma detection. When the CCD camera gain was varied from 1 to 103 , the plasma dimensions, measured at the lowest detectable part of the plasma plume luminosity profile, were determined to increase by a factor of 2. The plasma dimensions measured at FWHM of plasma plume luminosity profile were found to increase by 8%. The measurements carried out with a luminosity profile of 10% of the maximum luminosity demonstrated a 6% increase in the plasma dimensions. It was this luminosity level that was chosen for plasma dimension measurements in our experiments.

0,26 0,22 0,18 0,14 0,1

2 Experimental results

0

Figure 1 presents the dependence of crater depth on the laser pulse number for different laser pulse durations. As

Ablation depth (µm) 2 4 6

6

8

10

Fig. 2. Evolution of copper crater depth per pulse with laser pulse duration

0,07 ps 0,46 ps 0,8 ps 2 ps 10 ps 0,14 ps

100

D (µm)

50

0 0

4

Pulse duration (ps)

150 10 ps 2 ps 0,8 ps 0,46 ps 0,14 ps 0,07 ps

2

5

10

15

20

Number of laser pulses Fig. 1. Evolution of copper crater depth with laser pulse number for different laser pulse durations

0 0

10

20

30

Delay time (ns)

40

50

Fig. 3. Temporal evolution of copper plasma longitudinal dimension D for different laser pulse durations

S383

3 Discussion and conclusion The experimental studies of craters and plasma expansion at the interaction of the sharply focused laser beam with a copper target have demonstrated two distinctive features of laser beam-target interaction. The first belongs to a femtosecond regime, where the laser pulse terminates before the heated matter escapes from the surface. For pulse durations below 1 ps (1 ps is the characteristic time of electron-phonon interactions in solids), a two-temperature model [1] of laser beammetal interaction can be applied. Laser-heated electrons with a temperature several orders higher than the surface lattice temperature do not have sufficient time to exchange energy with the lattice during the laser pulse. Because of the high electron temperature gradient, the laser energy that has been absorbed by the surface is transmitted into the matter by electron diffusion. By the end of the laser pulse, we can detect an overheated layer on the surface with a depth significantly greater than the optical depth of the pulse penetration. Surface ablation and crater formation take place after the laser pulse. The laser energy is deposited in the matter without a laser beam-plasma interaction. The ablation efficiency does not depend on the laser pulse duration in the femtosecond regime. The second regime (with picosecond pulses) can be defined by a laser pulse duration that is long enough for surface evaporation during the laser pulse. Ablation takes place with laserplasma interaction. Part of the incident laser energy is absorbed in the near-surface plasma by Inverse Bremsstrahlung (IB) and by photoionization, and plasma shielding can occur. IB depends on the electron density, laser wavelength, and ion density. The photoionization depends on atom density, laser photon energy, and intensity I m for m-photon ionisation. In our experiments with 800 nm laser wavelength in the

ps regime, the absorption of laser energy by IB should dominate. The above effects result in a decrease in the amount of energy reaching the surface. Consequently, the ablation efficiency decreases as well. The studies by [2], analogous to our results, demonstrated that for fluences above 5 J/cm2 and for pulses longer than 5 ps the ablation efficiency decreased with increasing laser pulse duration. This effect was explained by hydrodynamic plasma expansion during the laser pulse, plasma shielding of the laser radiation, and an increase in heat conduction losses. Numerical simulations of laser ablation with short pulses [4] derived a laser ablation efficiency dependence similar to that obtained in our work. Thus we may conclude that, for our experiments with 800 nm laser pulses, the ablation efficiency is better for the cases without laser-plasma interactions; that is, for pulse durations shorter than 1 ps. Further experiments need to be carried out to evaluate the laser ablation efficiency for visible and UV short laser pulses.

References 1. S.I. Anisimov, B. Rethfeld: SPIE Proc. 3093, 192 (1996) 2. S. Nolte, C. Momma, H. Jacobs, A. Tünnermann, B.N. Chichkov, B. Wellegehausen, H. Welling: J. Opt. Soc. Am. B 14, 2716 (1997) 3. D. von der Linde, K. Sokolowski- Tinten, J. Bialkowski: Appl. Surf. Sci. 109-110, 1 (1997) 4. A.M. Rubenchik, M.D. Feit, M.D. Perry, J.T. Larsen: Appl. Surf. Sci. 127-129, 193 (1998) 5. M. Pessot, P. Maine, G. Mourou: Opt. Commun. 62, 419 (1987) 6. J.C. Diels, J.J. Fontaine, I.C. Michael, F. Simoni: Appl. Opt. 24, 1270 (1985) 7. G. Albrecht, A. Antonetti, G. Mourou: Opt. Commun. 40, 59 (1981)

Lihat lebih banyak...

Comentarios

Copyright © 2017 DATOSPDF Inc.