 |
| |
 |
| |
 |
| |
 |
|
| Copyright 2010 - MMendes Tech - All rights reserved |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
CHAPTER 5
STUDY OF THE INFLUENCE OF THE WORKING ATMOSPHERE ON THE MICROMACHINING
OF AL2O3-TiC CERAMICS
The effects of the ambient atmosphere on the ablation of Al2O3-TiC using nanosecond duration KrF (l=248 nm) laser pulses are studied. The influence of the gas type and pressure on the ablation rate and surface topography is investigated, and the observed changes are correlated with the ablation plume dynamics. The mechanisms of formation of particles during ablation are studied.
5.1 INTRODUCTION
Excimer laser micromachining induces extensive material removal from the near-surface region of target. During the pulse duration (typically between 10 to 60 ns), the laser radiation interacts simultaneously with the target and the ablated material, which may cause a drastic attenuation of the laser energy incident on the target [1-4]. Several mechanisms have been proposed to explain the attenuation of radiation:
- Inverse Bremsstrahlung (IB) heating of electrons through electron-neutral (EN) or electron-ion collisions eventually leading to avalanche breakdown [5, 6]. Inverse Bremsstrahlung involves the absorption of photons by free electrons, which gain energy from the laser beam during collisions with neutral and ionised atoms, thus promoting vapour ionisation and excitation through electron collisions with excited- and ground-state neutrals. In the IB process the probability of photon absorption due to electron ion collisions is much greater than that of scattering by neutral atoms [7]. However, a large density of neutrals in the initial vapour can significantly increase this last contribution [7]. The absorption of radiation by inverse Bremsstrahlung may be described by the absorption coefficient aIB (cm-1):
|
|
where Ne (cm-1) is the electron density and sIB (cm2) represents the total cross section of the process, with the contribution of both neutral (sIB,neu) and ion (sIB,ion) scattering included. The IB cross-sections are given [6, 8], respectively by
|
|
|
|
|
(5.1.3) |
where l is the laser wavelength (cm), Te the electron temperature (eV), hn the photon energy (eV), scoll the cross section for electron-neutral collisions (cm2), Z the ionic charge, and Ni and Nn are the ionised and neutral atoms number densities (cm-3), respectively. The cross section (sIB) wavelength scaling suggests that heating by this mechanism is relatively unimportant at short UV wavelengths [1, 4] for the modest electron densities initially existing in thermally ionised plumes. Despite the IB wavelength dependence, there are claims that IB can in fact be the predominant absorption mechanism for UV radiation [9, 10].
- Photo-ionisation from the ground state of neutral species or from thermally populated excited electronic states if the photon energy is high enough. Several studies have drawn attention to the importance of direct photo-ionisation of excited neutrals in UV laser ablation, mainly for metallic targets [1, 3, 7, 11]. The absorption coefficient of photo-ionisation is given by [6]:
|
|
where Nn and snn are the particle concentration and the cross-section at the energy level n, respectively. N1 is the atom concentration at the ground state, IH the ionisation potential, n the electronic energy level in the atomic structure, n the frequency of the photon, kB the Boltzmann’s constant and h Planck’s constant. The summation is performed over the energy levels where the electron-binding energy is less than the photon energy. Multiple photon absorption or photo-ionisation of ions are much less probable, and are not considered in the expression above [6]. The electrons generated by the photo-ionisation process are not in thermal equilibrium, and recombination reduces the concentration of electrons and ions in the plume. When the recombination time constant is comparable to or longer than the duration of the laser pulse, the electrons generated by the photo-ionisation process also contribute to photon-electron inverse Bremsstrahlung absorption [7, 11].
- Scattering
and absorption of radiation by
particles in the plume. Pulsed laser ablation results in the production of
particles which typically have two widely different sizes: nanoparticles [12,13], called by some authors debris [14, 15], with diameters varying from a few to several tens of
nanometers, and particulates with diameters up to several micrometers [16, 17]. Scattering [4,
18, 19] and/or absorption [20] of laser light by particles can cause a drastic
reduction in the radiation intensity transmitted through the plume. Using Mie
scattering theory [21], under the assumption that each photon is scattered
only once in the interaction zone by spherical material clusters with complex
refractive index 
and with a radius
a that is small compared with the
incident wavelength, Rayleigh approximation may be used. Under such
approximation, the attenuation coefficient aMie is described by
|
|
with Ncl being the cluster concentration and Cabs and Csca the cross sections for absorption and scattering respectively. The cross sections are described by [21]
|
and
|
|
with
|
|
The particles are
considered much smaller than the radiation wavelength when 
[21].
The formation of ultrafine debris has been explained by condensation of particles in the plume [15, 22], spallation [23] or instabilities at the liquid-gas interface [24]. The production of larger particles has been explained by phase-explosion [25, 26], the effect of the recoil pressure wave [27], hydrodynamic effects [28], exfoliation [14], or surface instabilities [24]. Redeposition of both kinds of particles limits surface finish and the achievable dimensional accuracy in micromachining [29, 30], while larger particulates destroy the uniformity of films deposited by PLD [16]. On the other hand, the formation of nanometer-sized particles by laser ablation can be put to advantage to produce nanocrystalline thin films and powders with interesting mechanical, magnetic or electric properties [31]. Ultrafine semiconductor particles produced by pulsed laser ablation have also received considerable attention because quantum confinement modifies the bulk band structure, leading to different radiative transitions with narrower linewidth emission [32]. The need to avoid particle formation or to control the characteristics of the particles produced by pulsed laser ablation has led to intense study of the particle formation mechanisms, as reviewed by Chen [16] and Bäuerle [17]. However, controversy remains concerning the relevance of the various mechanisms proposed.
The efficiency of the radiation extinction mechanisms is closely related to the density of species in the plume [33]. Iida [34] investigated the influence of the atmosphere (composition and pressure) on the laser-induced plasma, when processing various metal and ceramic targets with a Q-switched Nd:YAG laser (l=1.06 mm, t=10 ns). When processing aluminium, the estimated electron density 100-300 ns after the laser pulse was larger by an order of magnitude in Ar than in He atmospheres (760 torr: Ar 3.1´1018 cm-3, He 4.8´1017 cm-3; 100 torr: Ar 1.2´1018 cm-3, He 1.3´1017 cm-3). Similar results were observed by Aguilera et al. [35] using a 4.5 ns Nd:YAG laser (l=1.06 mm) to process low alloy steel samples (Figure 5.1.1), and by Schittenhelm et al. [4] using a KrF laser (l=248 nm) to process aluminium (Figure 5.1.2).
|
a) |
b) |
|
Figure 5.1.1: Steel laser produced plasma generated in air, argon and helium at atmospheric pressure (l=1.06 mm, t=4.5 ns) [35]. Temporal evolution of the a) Temperature and b) Electron densities. |
|
|
Figure 5.1.2: Three dimensional plot of the electron density distribution in argon, air, and helium at 1 bar, 36 ns after beginning of the excimer pulse (l=248 nm) when processing aluminium at 37 J/cm2 [36]. |
A number of studies have shown that increasing the pressure of the ambient gas leads to an increase of the confinement of the ablation plume [6]. Chu et al. [37] studied the ablation of titanium using a KrF excimer laser and fluence varying from 4 to 8 J/cm2 in an argon atmosphere with pressures up to 1 mbar. They verified that the background pressure limits plume expansion and consequently enhances collisions within the plume, as demonstrated by the decrease of the velocity of the neutral titanium atoms and the overall increase of emission intensity. Hermann et al. [38] performed time- and space- resolved plasma spectroscopy during XeCl (l=308 nm) excimer laser ablation of Ti in a low pressure N2 atmosphere. They observed that the decay rates of the plasma electronic density Ne and temperature Te are reduced with an increase of the ambient pressure P0, due to enhanced elastic and inelastic collisions between metal vapour and ambient gas species. The plasma density decreases slower and, as a result of inelastic collisions, the kinetic energy is partially transformed in excitation energy. During the early plasma phase (within 200 ns after the laser pulse), the authors observed that for P0 > 1 mbar both Ne and Te increased with increasing laser intensity, although at different rates. For ambient pressures P0 £ 1 mbar no influence on the values of Ne and Te was detected. They concluded that under these conditions the interaction laser radiation-material is not influenced by the presence of the ambient gas.
Iida [34] also investigated the influence of the gas type and pressure on the removal rate of different metal and ceramic targets using a Q-switched Nd:YAG laser (l=1.06 mm, t=10 ns). The author observed that 1) the amount of material vaporised is larger when processing in He than in Ar, and decreases with increasing pressure, which was related to an increased shielding effect due to inverse Bremsstrahlung absorption in the Ar denser plume; 2) the removal rate of ceramics is less affected by the ambient gas pressure than metals. This was explained by differences in the sample characteristics, which influence the beginning of the plasma generation process by providing the initial electrons: 1) metals have higher optical absorption coefficients than ceramics at this wavelength, and so higher absorbed power per unit volume is reached in metal targets, and as a result breakdown occurs more easily for metals than for ceramics [5]; 2) ceramics usually have higher melting and boiling points than metals, and so more laser power is needed to raise the sample surface sufficiently high for vaporisation [17]; 3) multiphoton ionisation may become important for dielectrics [39]. Russo et al. [10] investigated the influence of noble gases in the ablation rate of brass when processing using a XeCl excimer laser. The ablation rate is highest in He and decreases for Ne, Ar, Kr and Xe atmospheres. Russo suggested that “plasma shielding” occurs in the plume through inverse Bremsstrahlung, limiting the available energy for ablation, and that the ionisation potential of the ambient gas changes the power density at which laser energy is coupled to the laser induced plasma through IB instead of the target surface. Conversely, Arnold [40] studied the KrF laser ablation of molybdenum, copper and tantal in different ambient atmospheres at 1 bar using a laser fluence of 15 J/cm2, and observed no dependence of the ablation rate on the atmosphere composition, with practically no variation between helium and argon. These results show that the effect of the atmosphere (composition and pressure) on laser ablation efficiency is material-specific, even within the same class of materials.
The atmosphere may affect considerably the ablation process and induce chemical changes on the irradiated surface [17]. Cao et al. [41] studied the effects of air and Ar-4% H2 atmospheres at atmospheric pressure on the surface modification of alumina. They used polycrystalline alumina substrates of nominal purity 99.6 %, which were irradiated using XeCl laser pulses (l=308 nm) and fluence in the range 1-4 J/cm2. In an Ar-4% H2 atmosphere they found that an amorphous layer was formed at a fluence between 1 and 1.3 J/cm2, while at 1.6 J/cm2 or higher the melted alumina solidified as a crystalline phase. The microstructure consisted of cells and small aluminium particles distributed throughout the resolidified alumina. Specimens irradiated in air at 3 J/cm2 showed the same cellular structure as those irradiated in the Ar-4% H2 atmosphere, but no particles were detected in the solidified region. EDS analysis showed that the O/Al concentration ratio is higher in irradiated areas than in unirradiated areas. These results demonstrate the relevance of the atmosphere during laser processing of alumina, in opposition to the studies by Capelli [42] and Sciti [43]. Cappelli et al. [42] studied the surface modifications induced in technical grade sintered and polished polycrystalline alumina substrates of nominal purity ~90% (a-Al2O3, ~10 % Al6Si2O16 and traces of Al2TiO5) after irradiation by an ArF excimer laser (l=193 nm) in vacuum (~10-3 Pa), in argon (10 sccm flow) and in pure oxygen (100 sccm flow). XPS analysis showed no evidence of metallic aluminium, while no significant variation of the atomic percentage of aluminium and oxygen was detected through analysis of the XPS depth profile composition. Sciti et al. [43] irradiated 90% pure a-Al2O3 using a KrF excimer laser (l=248 nm) and verified that surface morphology and composition do not depend on the gas (air or argon).
5.2 EXPERIMENTAL
The experiments were performed on substrates of an Al2O3-TiC composite ceramic consisting of 70 wt. % of alumina and 30 wt. % of titanium carbide. Processing was carried out by mask projection using the micromachining system described in chapter 4. The laser was operated with a KrF gas mixture (l=248 nm, t=30 ns FWHM). The influence of the spot diameter on the ablation rate was investigated with constant laser pulse energy, by varying the diameter of a circular aperture, which selected a homogeneous part of the laser beam. The experiments were performed in ambient air or under an atmosphere of He, Ne, N2, Ar, or Kr. Controlled atmosphere experiments were performed using a stainless steel vacuum chamber (Figure 5.2.1), which was filled with the required gas up to 1 bar and evacuated to ~10-6 bar at least 3 times before setting the predefined gas pressure for processing (varied between 10-5 and 1 bar). For comparison purposes, a few experiments were carried out in N2+40% Cl2 atmosphere at 1 and 10 mbar. In addition, some experiments were performed on TiC samples (99.5 wt. %) in air at ~10-4 bar.
|
|
|
|
Figure 5.2.1: Set-up for processing in controlled atmosphere, showing the vacuum chamber used. |
|
The shock wave generated when the ablated material expands into the atmosphere was studied by a probe-beam deflection technique [44]. In probe-beam deflection experiments, variations in the refractive index due to the expanding shock wave cause the deflection of a probe laser beam away from a detector, and consequently change the detected intensity. In the present work, a probing He-Ne laser beam parallel to the surface of the sample and crossing the interaction zone was used. Detection was performed by an intensity-sensitive avalanche photo-diode (APD), located 0.8 m away from the irradiated area. By incrementally increasing the distance between the probing beam and the sample surface, and determining the variation of the arrival time of the shock wave as a function of the distance to the target, the velocity of the laser-induced shock wave was calculated.
The optical emission from the plume was analysed using a Echelle spectrometer (Mechelle 7500, Multichannel Instruments), with a theoretical resolution of l/Dl = 7.5 ´ 103, constant over the entire spectral range 180-1000 nm. The spectrometer was coupled to an ICCD (Intensified Charged Couple Device) camera (PCO Dicam Pro) with minimum obturation time »3 ns and 12 bits resolution (Figure 5.2.2). This system was also used to analyse the evolution of the laser intensity reflected from the target. The laser beam was incident at 45° to the surface normal, while the reflected laser light was collected by an optical fibre, located at an angle of 90° to the beam, and its intensity monitored by a spectrometer at l=248 nm.
The ICCD camera (Dicam Pro, PCO) was also used for ultrafast imaging of the ablation plume by coupling it to a proper objective. Images of the plume were also recorded using a CCD-videocamera with a fast electronic shutter and a spectral range of 400-1000 nm.
|
Figure
5.2.2: Photo of the spectrometer coupled to
the ICCD detector. |
After laser processing, the samples were observed by optical microscopy, scanning electron microscopy (SEM) and atomic force microscopy (AFM). Chemical analysis was performed by energy dispersive spectroscopy (EDS). The radial extension of the area where redeposited material accumulates was taken as the maximum distance from the irradiation centre. The volume of accumulated material was calculated from SEM micrographs and surface topography profiles. Chemical analysis of the redeposited material was performed by EDS in material collected on a copper sheet located near the processed area.
FOR RESULTS, DISCUSSION, CONCLUSIONS AND BIBLIOGRAPHY CHECK THE THESIS MAIN PAGE