Microstructuring of glass by laser irradiation: applications on microoptics and microfluidics

Daniel Nieto García Área de Óptica, Departamento de Física Aplicada Universidade de Santiago de Compostela -7 60 60 120 180 240 180 120 0 0 Z (µm) Y (µm) X (µm) -1 -10 -4 -7 MICROSTRUCTURING OF GLASS BY LASER IRRADIATION: APPLICATIONS ON MICROOPTICS AND MICROFLUIDICS MICROESTRUCTURADO DE VIDRIO MEDIANTE IRRADIACIÓN LASER: APLICACIONES EN MICROOPTICA Y MICROFLUÍDICA) DOCTORAL THESIS -7 60 60 120 180 240 180 120 0 0 Z (µm) Y (µm) X (µm) -1 -10 -4 -7 Daniel Nieto García Área de Óptica, Departamento de Física Aplicada Universidade de Santiago de Compostela ( Dña. María Teresa Flores Arias y D. Justo Arines Piferrer, Profesora Titular de Universidad e Investigador “Isidro Parga Pondal”, respectivamente, del Área de Óptica del Departamento de Física Aplicada de la Universidade de Santiago de Compostela. INFORMAN que la presente memoria, titulada “ Microstructuring of glass by laser irradiation: applications on microoptics and microfluidics” ha sido realizada por D. Daniel Nieto García bajo su dirección y constituye la Tesis que presenta para optar al Grado de Doctor en Fotónica e Tecnoloxías do laser Santiago de Compostela, 1 de Junio de 2012 Vº. Bº Director de la Tesis Vº. Bº Director de la Tesis María Teresa Flores Arias Justo Arines Piferrer El Doctorando Daniel Nieto García This work has been funded by the Ministerio de Ciencia e Innovación (MICINN) under the projects TEC2006 10496, CEN 2007-2014, SURFALUX SOL-00030930 and MAT2010-18519; by Xunta de Galicia/FEDER through the project INCITE08PXIB206013PR. a Mar y a Izan Summary The aim of this thesis is to develop a method for fabricating microlens and microfluidic microchannels on soda-lime glass. The method consists in a combination of the laser direct write technique for fabricating the promoting glass structures, a wet chemical etching to remove the imperfections generated during laser ablation and a thermal treatment for reshaping and or improving of the optical and morphological qualities of the generated microlens and microfluidic microchannels. The laser direct write technique used for glass microstructuring consists of a pulsed laser combined with a galvanometer system for addressing the output laser beam. Using this configuration cylindrical post and microchannels on sodalime glass are fabricated. A later wet chemical etching technique (HF) is used successfully to carry out the surface smoothing and eliminate secondary structures generated during lasing the soda-lime glass. The mechanism to modify the surface roughness and shape by etching is also studied. Finally, a thermal treatment is used to modify the shape and the roughness of the element. Thanks to the thermal reflowing of glass it was possible to obtain microlens and microchannels in different configurations with a roughness of 7.35 nm, which is two orders of magnitude below the roughness after laser microstruturing. It was also investigated the influence of temperature, heating time and the mechanism of thermal displacement on the shape of the fabricated elements. In parallel to the fabrication process, it was investigated the interaction between a laser beam and the glass materials, presenting an overview of the ablative process on soda- CHAPTER 1 MICROOPTICS FABRICATION TECHNIQUES This chapter gives a general introduction to the role of glass materials and its applications on micro-optics and microfluidics technologies. Besides, describes the commonly used methods for microstruturing glass materials, focusing on the role of lasers and its different approaches. Microoptics fabrication techniques CHAPTER 1 2 1.1 Introduccion Glass materials due to their inertness and other thermo-mechanical attributes are suitable for many micro and nanotechnological applications. Micro-optics and microfluidics demands high quality microstrutures on glass. The large and well developed glass micromachining toolbox provide the capability to obtain microstructures with high precision and repeatability. Common techniques used for microstructuring glass are wet etching, deep reactive ion etching, precision glass molding, polishing, griding and laser ablation. Laser ablation, because of its non-contact nature, allows the micromachining and surface patterning of glass materials with minimal mechanical and thermal deformation. Due to the low cost, the ease to implement in industry, and the wide quantity of lasers available, this technique is extremely important for machining glass. 1.2 Importance of glass. Glasses are extremely useful and important optical materials, which have been widely used in optical instruments as windows or lenses. Due to their inertness and other thermo-mechanical attributes, glasses are suitable for many applications. Glasses are not crystalline solids, so do not exhibit the optical anisotropic that is characteristic of some crystals, and because glass has quite distinct properties from silicon, PMMA and metals, applications exist where only glass devices meet the requirements [Karthe1995]. Most glasses are composed of sand (Silica) mixed with other chemicals. By adding different chemicals to silica during the fusion process it is possible to fabricate different kinds of glass that differ in the refraction index and the wavelength range for transmission. It can be expected that, since silica is a glass, it presents Microoptics fabrication techniques CHAPTER 1 3 the same characteristic as insulators. Optically, they are transparent in the visible region, absorb in the UV due to the electronic transitions, and in the IR due to vibrational absorption [Sugioka2004]. The main advantages of glass derive from its specific characteristic including properties such as that it is natural, inexpensive and highly aesthetic. These make it suitable for manufacturing on sustainable and healthy material. For many applications in clinical medicine, biology and chemistry, glass is the preferred material. Some of these applications involve the microstructuring of glass [Coughlan2008]. Micro-optics: micro-optics play an important role in various industrial production processes, like micro-electronics, laser micro-machining, materials processing, optical inspection, machine vision and precision metrology. Micro-optics elements are additionally found in the medical industry, analytical systems in bio-photonics and optical sensors. Various applications and their subsequent requirements with respect to cost structure, accuracy, quality and lifetime, caused different manufacturing technologies [Flores-Arias2009, Sohn2005]. Traditionally, three different kinds of materials are used for micro-optics: glass, polymers and crystal. Of course polymer optics is the most cost-effective solution, however, they have some limitations due to the reduced number of materials with different refractive index, limited stability at higher temperatures, burning at high laser powers, transparency and degradation under UV-light. These limits can be overcome by using glass. Microfluidics: Microfluidics refers to a set of technologies that control the flow of minute amounts of liquids or gases in a miniaturized system. The use of Microoptics fabrication techniques CHAPTER 1 4 microfluidic devices is making rapid inroads in the modern analytical laboratory, primarily because of their small physical footprint, speed and efficiency of chemical separations, and reduced reagent consumption. Traditionally, lab-on-achip devices have been manufactured in silica due to its well understood surface chemistry and favourable micromachining techniques that are ubiquitous in the microelectronics industry. Recently researchers have begun to utilize devices fabricated from polymer substrates as an alternative to glass, although in clinical medicine, biology and chemistry, glass may still be the preferred materials. Microchannels fabricated on glass have a growing importance in the miniaturization of microfluidic devices for chemical and biological micrototal analysis system [Cheng2004, Malalahalli2004, Erickson2004]. 1.3 Microstructuring of glass. The large and well developed glass micromachining methods provide the capability to obtain microstructures with high precision and repeatability. Common techniques used for microstructuring glass are laser ablation, etching, precision glass molding and mechanical micromachinig. Laser micromachining offers a single-step method for direct writing of microchannels in glass. Using this laser ablation technique, it is possible to fabricate geometries with variable depth and high aspect ratio that cannot be achieved through traditional microlithographic techniques. The mechanical rigidity, chemical resistance, and low permeability properties of glass, combined with their optical transparency, make them a good choice for many demanding lab-on-a-chip applications. Although some techniques have been widely used for structuring glass most of them are complex, high in cost and come with significant contamination risk. Table 1.1 shows the common methods used for microstructuring glass materials. Microoptics fabrication techniques CHAPTER 1 5 Table 1.1 Commonly methods used for microstructuring glass. Technology Substrate configuration Material Tool/Master/Mask Laser-direct write Substrate based Glass None Laser-Indirect write Substrate based Glass Absorption source Etching Wafer based Glass Mask for optical lithography Precision glass molding Substrate based Special thermal Glass Mold pre-form Mechanical micromachining: griding and polishing Single optics Glass None 1.3.1 Direct Laser ablation of glass. Laser has become increasingly important in recent years for many fields, including micro-optics, micro-electronics, micro-biology, and micro-chemistry. High micromachining quality with ns-pulse and fs-pulse lasers was demonstrated for direct ablation of dielectrics [Liu1997]. Laser ablation, because of its non-contact nature, allows the micromachining and surface patterning of glass materials with minimal mechanical and thermal deformation. Due to the low cost, the ease implementation in industry, and the wide quantity of lasers available, this technique is extremely important for machining glass. For transparent materials, in the visible spectral range, laser ablation should ideally be performed with ultraviolet radiation because of the linear optical Microoptics fabrication techniques CHAPTER 1 6 absorption in this wavelength range. Alternatively, non-linear coupling of ultrahigh-intensity laser pulses in the near-IR range with sub-picosecond duration may show advantages. The interaction between a laser beam and a material is determined by: (1) laser characteristics (the wavelength, fluence or energy density, the pulse duration, repetition rate and pulse energy); (2) the properties of the material (absorption characteristics, thermal relaxation,) which is governed by the composition and structure of the material. From the earliest work on laser interactions with materials, direct-write processes have been important and relevant techniques to modify, add, and subtract materials for a wide variety of systems and applications. In general, direct-write processing refers to any technique that is able to create a pattern on a surface or volume in a serial or “spot-by-spot” fashion. This is in contrast to lithography, stamping, directed self-assembly, or other patterning approaches that require masks or pre-existing patterns [Hirai2003, Wensink2000]. Laser direct-write subtraction (LDWS) is the most common type of laser directwrite. In general, this entails processes that result in photochemical, photothermal, or photophysical ablation on a substrate or target surface, directly leading to the features of interest. Common processes include laser scribing, cutting, drilling, or etching to produce relief structures or holes in materials in ambient or controlled atmospheres. The fundamental interactions leading to material removal can be thermal or athermal, depending primarily on the material characteristics and the pulse duration of the laser [Li2001, Mannion2004]. On laser direct-write modification (LDWM) the incident laser energy is usually not sufficient to cause ablative effects but is sufficient to cause a permanent change in the material properties. Typically, these processes rely on thermal modifications that cause a structural or chemical change in the material. A Microoptics fabrication techniques CHAPTER 1 7 common example of such processes is the rewritable compact disc, in which a diode laser induces a phase transition between crystalline and amorphous material. Many LDWM applications require a specific optical response in the material of interest beyond simple thermal effects. Optically induced defects or changes in mechanical properties can lead to many non-ablative material modifications [Cheng2004]. Laser Direct-Write Addition (LDWA) is perhaps the most recent of the laser directwrite processes. In this technique, material is added to a substrate using various laser-induced processes. Many techniques are derived from laser induced forward transfer (LIFT), where a sacrificial substrate of solid metal is positioned in close proximity to a second substrate to receive the removed material, the incident laser is absorbed by the material of interest, causing local evaporation. This vapor is propelled toward the waiting substrate, where it recondenses as an individual three-dimensional pixel, or voxel, of solid material [Piqué2006]. 1.3.2 In-direct laser ablation of glass. Since the ablation of transparent materials using IR wavelength is complicated due to low absorption, it can be easily understandable that some kind of element placed at the surface or backside of the transparent material can be able to confine the impinging laser beam which will lead to transfer energy from the interaction laser-absorber layer to the transparent material. There are several indirect laser methods, which are suitable processing of nonabsorbing targets. One of them is laser-induced plasma-assisted ablation (LIPAA). Another group of indirect methods are based on an absorbing material contacting the backside of the transparent target. The most common techniques that frequently use absorbers are: condensed vapour in the laser etching at a surface Microoptics fabrication techniques CHAPTER 1 8 adsorbed layer (LESAL), liquids in laser-induced backside wet etching (LIBWE) or solid thin films in the case of laser-induced backside dry etching (LIBDE). In the LIPAA, the substrate must be transparent to the wavelength of the laser beam, so the laser beam goes through the substrate first, and then irradiates a metal target placed behind the substrate. At laser fluence below the damage threshold of the substrate and above the ablation threshold of the target, the plasma generated from the target surface propagates to the rear surface of the substrate at high speed. Strong interactions among the laser beam, the plasma, and the substrate take place, resulting in high-quality ablation with neither cracks nor severe distortions at the rear side of the substrate [Sugioka2004, Hanada2004]. A similar configuration is needed in case of LESAL. The substrate must be transparent for the wavelength of the laser beam, so the radiation goes through the substrate first, and then irradiates the absorber layer deposited at the rear side of the substrate. The absorber layer is usually deposited by vapour deposition. At laser fluence below the damage threshold of the substrate and above the ablation threshold of the target, the plasma generated from the interaction of the laser with the absorber layer transfers the energy to the substrate which leads to the ablation process [Ihlemann2008, Hopp2009, Zimmer2008]. LIBDE follows a configuration and mechanism like that of LESAL, although differs on using a solid thin film at the rear side of the transparent materials to be etched [Hopp2009]. On LIBWE the laser beam passes through a transparent substrate and is absorbed by a dye solution located at the rear surface of the substrate [Kopitkovas2003, Kopitkovas2007, Smausz2007, Huang2007]. The principle of LIBWE is based on the generation of intense temperature and pressure increase at the thin interface between the material and liquid. Due to strong laser absorption Microoptics fabrication techniques CHAPTER 1 9 by the dye solution, the laser energy is confined into a thin layer at the interface between the transparent substrate and the solution. The photo-activated layer interacts with the surface of the substrate, resulting in etching. The major difference with other techniques which are used for structuring of transparent materials is the application of conventional excimer lasers and an organic liquid which strongly absorbs the laser radiation. 1.3.3 Etching. Etching is used in microfabrication to chemically remove layers from the surface of a wafer during manufacturing. Wet etching technique (HF etching) is commonly used to create channels and structures in glass [Haixin2009, Grosse2001, Steingoetter2005]. In the HF technique, etching is used to chemically remove the glass materials from the surface preciously protected using a mask. With this technique the amount of channels, reservoirs, mixing and reaction chambers is irrelevant for the cost of the process. Wet etching of glass is an isotropic etching technique. This means that the width of the channels is more than twice the depth of it, and that the corners are rounded. The bottom of the channel stays smooth and optically transparent. Another etching method for machining glass is Reactive-ion etching (RIE). It uses chemically reactive plasma to remove material deposited on wafers. The plasma is generated under low pressure (vacuum) by an electromagnetic field. High-energy ions from the plasma attack the wafer surface and react with it. Plasma is initiated in the system by applying a strong electromagnetic field (in the radio frequency range) to the wafer platter. The oscillating electric field ionizes the gas molecules by stripping them of electrons, creating the plasma [Thienot2006, Righini1991]. Because of the large voltage difference, negative ions tend to drift toward the Laser-glass ablation mechanism CHAPTER 2 16 2.1 Introduction During laser ablation, the energy that reaches the sample is absorbed, transmitted and reflected depending on the properties of the materials. The laser parameters, wavelength, pulse duration, repetition rate and the energy density of fluencies applied to the samples also determines the response of the material. The laser beam is absorbed by the sample through electronic and vibration excitation of the atoms, if the electric field exceeds a critical value; it results in a rapid ionization and the formation of plasma. When the density of free electrons is above the plasma threshold, irreversible material breakdown occurs and ablation begins. The temperature increases with the number of photons reaching the sample, which also increase the fraction of light absorbed. At this time, the electrons absorb energy by collisions with ions and are heated to higher temperatures. The electrons also transfer energy to the lattice, which contributes to material heating. The amount of energy transfered to the lattice during the laser pulse depends on the pulse duration and on the energy coupling coefficient. For pulse durations longer than a few tens of picoseconds, energy goes from the electrons to the lattice on the time scale of the pulse duration. In this case the damage occurs via heat deposition resulting in melting and boiling, and consequently an expansion and expulsion of material occurs [Yamamoto2010, Huang2010]]. On the other hand, for pulses shorter than a few picoseconds, the time scale for absorption is shorter than for energy transfer to the lattice, which decouples the absorption and the heating process of the lattice, which result in a direct vaporization of the material from the surface. Laser-glass ablation mechanism CHAPTER 2 17 2.2 Laser Ablation mechanism Laser induced breakdown is a process where a normally transparent material is first transformed into an absorber plasma by the strong laser pulse. After that, the plasma absorbss the laser energy and it causes an heating that leads to an irreversible damage at the surface material. A conceptual framework for a physical model of laser ablation based on electronic processes incorporates [Miller2002]: - The primary interaction mechanism of photons with solids surfaces, - Electronic, vibrational and configurational electron-lattice relaxation, - The evolution of defect and other characteristic vacancy cluster, - Secondary energy-dissipation processes that produce the catastrophic nonlinear response typical of laser ablation. The pulse duration and so the material response time is an important factor that will determine the kind of mechanism that leads to ablation of material. Figure 2 Diagram of valence and conduction band for insulators, semiconductors and metal. Laser-glass ablation mechanism CHAPTER 2 18 The energy bands each correspond to a large number of discrete quantum states, and most of the states with low energy are full, up to a particular band called the valence band. The easiness with which electrons in an insulator can be excited from the valence band to the conduction band depends on the energy gap determining the wavelength range over which interband absorption takes place (fig.2.1). The absorption of light, which leads to the ablation of transparent materials, comes from linear and nonlinear effects. Important characteristic of laser parameters such pulse durations and photon energy are involved in the kind mechanism that predominates. While using IR wavelenghts and long pulses, avalanche effects is the responsible mechanism, with short pulse durations multiphoton absorption takes place [Sze2001, Fox2010]. As pointed above, the absorption of light, which leads to the ablation of materials, comes from a photophysical response which involves inter-band, intra-band and impurities effects (Fig.2.2). Figure 2.2 Diagram of the mechanism of absorption of light on transparent materials through a) single and multi-photon inter-band absorption and b) impurity-based absorption. Laser-glass ablation mechanism CHAPTER 2 19 At the initial stages of pulsed laser interaction with dielectrics, the process that contributes to the formation of free electrons are multiphoton absorption, absorption at lattice and surface defects, nonlinear absorption and chemical decomposition. At high intensity threshold a free electrons gas in the conduction band of the dielectric is formed, which leads to a strong absorption. Classically, a free-electron generation in dielectrics is given by a single rate equation describing the increase of the total free electron density inherent in the conduction band ntotal, eq. (2.1) [Rethfeld2006], (2.1) This equation combines the probability of photoionization , directly depending on the amplitude of the electric laser field , with the impact ionization, assumed to depend on the total free electron density. Due to photoionization, electrons are shifted from the valence band into the conduction band, [Keldysh1965]. In contrast electro-electron impact ionization is caused by a free electron already existing in the conduction band. If its kinetic energy is large enough, it may transfer part of it to an electron in the valence band, such that the latter is enabled to overcome the ionization potencial [Keldysh1960, Bloemberger1964]. The avalanche coefficient depends on the effective energy gain of the free electron in the electric laser field and can be estimated by eq.(2.2) (2.2) where, is the probability of one intraband photon absorption, the photon energy of the laser light, and the critical energy of the impact Laser-glass ablation mechanism CHAPTER 2 20 ionization, wich is on the order of Egap, the band gap between the valence band and the conduction band. Eq. 2.1-2.2 was proposed and verified for laser pulses in the nanosecond regime[Smith1978, Jones1989] Experimental studies applying eq. 2.1, have lead to contradictory results, refuting [Stuard1996, Apostolova2000] or emphasizing [Kaiser2000, Vatsa2002, Sudrie2002] the importance of the electron avalanche in picosecond and below, which has introduces doubts about the application of eq. 2.1 to ultrashort times scales [Manenkov1986, Bityurin2003]. One basic assumption of eq. 2.1 is that impact ionization depends directly on the total density of free electrons. However, the energy of a particular electron plays an important role, since the photoionization generates electrons with low kinetic energy in the conduction band while impact ionization requires electrons of high kinetic energies. The additional energy is absorbed from the laser light by intraband absorption. If this absorption process takes time comparable to the laser pulse duration, it can be appreciated that eq. (2.1) is oversimplified. Multiple rate equitation (MRE) [Rethfel2004] provides a possibility of clarifying these controversies within the framework of a unified approach, valid on a broad range of time scales. On ultrashort time scales the shape of the electron distribution in the conduction band may change on time, thus energy-averaged total electron density ntotal is not the adequate parameter to describe the ionization process. Defining the density nk of electrons above , where will be identified with the number of photons necessary to reach , a modified rate equation from ref [Renthfelt2004] can be formulated as eq. (2.3) Laser-glass ablation mechanism CHAPTER 2 21 (2.3) The difference in the last term of eq. 2.3 compared with eq.(2.1) is substantial. While in eq. (2.1) the impact ionization is assumed to depend on the total freeelectron density, , eq. (2.3) considers the fact that only those electrons which bear sufficiently high energy may produce impact ionization. Figure 2.3 shows an illustration of the process providing changes in the density and the energy, respectively, of free electrons in the conduction band of a dielectric. Figure 2.3 Illustration of the process of impact ionization. [Rethfeld2006] When the density of electrons reaches the plasma threshold, irreversible material breakdown occurs and ablation takes place. After that the electrons transfer energy to the ions and the lattice, which results in heating of the material. The amount of energy transfer, hence the heating, during the laser pulse are conditioned by pulse duration and the energy coupling coefficient. Since the heating time is equivalent to the pulse width, different interactions times are involved in transferring the energy from the laser to the lattice of the material. The ultra short pulse laser-matter interaction mode corresponds to conditions when the electron-to-ion energy transfer time and the heat conduction time exceed significantly the pulse. With all types of materials, ablation is strongly n0n1nk-1 … … Photoionization Impactionization Sequencial 1pt ionization nk Ԑ 0Ԑ 1Ԑ k-1Ԑ k Ԑ crit Ԑ Conduction Band Valence Band ε0ε1εk-1 εk εcrit ε Laser-glass ablation mechanism CHAPTER 2 22 non-stationary and starts only after about 1 to 10ps. The electron-phonon/lattice coupling time is in the range of 1 ps and the overall relaxation time is in the range of 200 ps (fig. 2.4). Figure 2.4 Sketch of electron-phonon coupling time. The electron-to-lattice energy exchange time is, in a long-pulse ablation mode, of several orders of magnitude shorter than the pulse duration. By this reason the electrons and ions are in equilibrium, and ablation has a conventional character of thermal expansion. By contrast, for the short pulse interaction the electron-to-ion energy exchange time, as well as the heat conduction time, is much larger than the pulse duration, and the ions remain cold. Electrons can gain energy from the laser field in excess of the Fermi energy, and escape the target. The electric field of a charge separation pulls ions out of the target thus creating an efficient nonequilibrium mechanism of ablation. In conventional continuous wave and long pulse (nanosecond) laser processing, the mechanism that dominates the ablation process is the heating of the target material through the liquid phase to the vapor phase, resulting in expansion and expulsion of material. This is accompanied by heating and collateral damage to the surrounding area, the degree of which is determined by the rate of energy absorption and the rate of energy loss through thermal conduction in the τheating ≈τpulse τelec-phonon≈τpicosecond τrelajation≈200 picoseconds electrons coupling Equilibrium lattice electrons coupling lattice LaserLaser AblationAblation Nanosecond Femtosecond Laser-glass ablation mechanism CHAPTER 2 23 material. This collateral damage is a limiting factor when high precision ablation is required. On the other hand, due to their peak intensities, ultrashort (picosecond and femtosecond) pulses ablate material via the rapid creation of plasma that absorbs the incident energy resulting in direct vaporization of material from the surface. This produces negligible collateral heating and shock-wave damage (fig.2.5). Figure 2.5 Sketch of the different mechanism related with the pulse width. By decreasing the pulse widths, the laser field strength increases, which means short interaction time. The electrons achieves more energy, reaching a few to tens of electron volts, in contrast with the low energy of the lattice, which remain relatively cold. After the laser pulse is over, the electron–ion energy transfer will heat the ions to much higher temperatures compared to the long pulse cases and the material is vaporized directly without suffering a melting phase. Because of the short interaction time the heat affected zone is significantly smaller. Heatdiffusion Heat diffusion Shock waves Laser-glass ablation mechanism CHAPTER 2 24 After reaching the melting (long pulses) or vaporization (short pulses) temperatures, the material is damaged, which is also accompanied by acoustic waves and optical radiation. The energy is deposited in a surface layer and the thickness is given by the absorption or penetration skin depth , where is the absorption coefficient. The heat diffusion length during the pulse gives the heat penetration depth due to thermal conduction. This diffusion length is given by eq. 2.4. (2.4) where D is the heat diffusion coefficient and is the laser pulsewidth. For long pulses, > and the volume of the material heated by the laser pulse, hence the temperature, is determined by the heat diffusion length during the laser pulse. Therefore, for long pulses, the fluence breakdown threshold varies with laser pulsewidth as . As the laser pulsewidth decreases to a value such that < , the skin depth determines the heated volume during the laser pulse. It can be applied to transparent materials once significant absorption due to plasma generation occurs, and it has been used to explain the observed scaling of the breakdown threshold [Krzystof2005]. When using long pulses a large volume of material is heated and melted, the ablation and material removal is accomplished through melt expulsion by the vapor pressure and the recoil of the light pressure. In this case the heat diffusion is larger than the skin depth. Laser-glass ablation mechanism CHAPTER 2 25 In some micromachining applications with long pulse lasers the melt material generated after the ablation can lead to very irregular shapes. On the other hand, when using ultrashort pulses, the deposited laser energy is limited in a layer with a thickness , so the material pass from the liquid phase to the vapor phase with high kinetic energy. Plasma shielding is strongly diminished or even avoided. Thus, ultrashort pulses allow strong material excitation prior to the expansion of the plasma plume [Baüerle2006]. The heat affected zone tends to be smaller than the focus when applying a laser with fluence just at the ablation threshold. The material is removed away by direct vaporization from the surface. Most of this process takes place after the laser pulse is over. Although, the material is still heated, the thickness of the melt layer is small. As the heated material is evaporated the material turn into cold due to the step temperature gradient and because of the absorbed laser energy is carried away by evaporation. Because of this ablation and material removal using short pulses is more precise compared with long pulses case. Materials, instrumentation and characterization methods CHAPTER 3 32 Figure 3.3 Picosecond laser work station. Figure 3.4 Femtosecond laser work station. Materials, instrumentation and characterization methods CHAPTER 3 33 Table 3.3 Specifications of lasers. Laser parameters Femtosecond Picosecond Nanoseocond Laser sources: Diode Pumped Femtosecond based on Ytterbium crystal technology Diode Pumped Picosecond laser based on Thin Crystal Disk technology Nd:YVO4 Wavelength: 1030 nm 1030 nm 1064nm Pulse Duration: 500 fs <10 ps 8-200 ns Pulse Energy: 1mJ 125 µJ 105 µJ Average Power: up to 4W 50 W 20W Pulse Repetition Rate: Single Pulse to 300 kHz Single shot to400 kHz Single shot to - 200kHz M2 1.2 1.3 1.2 3.4 Thermal treatment. The thermal treatment was applied in two different ovens; (1) a Heraeus mufla oven; (2) and a VITA Vacumat (fig 3.5). Table 3.4 shows the operational specifications of both devices. a) b) Figure 3.5 a) Heareus mufla oven and b) VITA Vacumet mufla oven. Materials, instrumentation and characterization methods CHAPTER 3 34 Table 3.4 Operational specifications for the Mufla ovens. Heraeus mufla oven VITA Vacumet 6000M Temperature range 40 ºC1100 °C Adjustable fresh air supply Thermal stability ± 1 ºC ….±10 ºC Thermal speed 0.1 ºC/min….40 ºC/min Uniformity at 1000: ± 4 ºC Thermal Accuracy +/- 1 ºC Thermicon P digital program controller Temperature range 200 ºC1100 °C Vacuum setting options, pre-vacuum and main vacuum can be selected Temperature accuracy +/- 1 °C 3.5 Characterization methods The tools used for determining the topographical and compositional data were a confocal microscope and a scanning electron microscope, respectively. 3.5.1 Confocal microscope The Confocal microscopy is an optical imaging technique used to increase optical resolution and contrast of a micrograph by using point illumination and a spatial pinhole to eliminate out-of-focus light in specimens that are thicker than the focal plane. It enables the reconstruction of three-dimensional structures and therefore the conduction of topographical analysis. In this thesis we use the confocal microscope SENSOFAR 2300 (fig. 3.6) [Sensofar]. This instrument is a non-contact optical profiler which uses a dual-technology sensor head combining both confocal and interferometric techniques. Materials, instrumentation and characterization methods CHAPTER 3 35 Figure 3.6 Confocal microscope SENSOFAR 2300 plu The measurements were performer using a 20XEPI microscope objective, which was selected according to the resolution features and the requirements of the sample to analyze. Table 3.5 shows the confocal specifications in function of the selected objective. Table 3.5 Confocal objectives specifications. Confocal objectives 5XEPI 10XEPI 20XEPI 50XEPI 100XEPI Working Distance (mm) 23.5 17.3 4.5 1.0 1.0 Numerical aperture 0.15 0.3 0.45 0.8 0.9 Field of View (µm) 2546x1909 1273x955 637x437 253x190 123x92 Spatial Sampling(µm) 3.32 1.66 0.83 0.33 0.17 Maximun Slope 8.5º 14º 21º 42º 51º Repeatability Rms(nm) ˂100 ˂50 ˂20 ˂4 ˂3 Acquisition time Confocal image(s) Topography(µm/s) 0.5 50-800 0.5 25-400 0.5 12.5-200 0.5 2.5-40 0.5 2.5-40 3.5.2 Scanning electron microscope. A scanning electron microscope (SEM) is a type of electron microscope that images a sample by scanning it with a high-energy beam of electrons in a raster Materials, instrumentation and characterization methods CHAPTER 3 36 scan pattern. The electrons interact with the atoms that make up the sample producing signals that contain information about the sample's surface topography, composition, and other properties such as electrical conductivity. Energy-dispersive X-ray spectroscopy (EDS or EDX) is an analytical technique used for the elemental analysis or chemical characterization of a sample. It is a form of X-ray fluorescence spectroscopy which relies on the investigation of a sample through interactions between electromagnetic radiation and matter. It analyzes Xrays emitted by the matter in response to being excited by electrons. Its characterization capabilities are due in large part to the fundamental principle that each element has a unique atomic structure; the X-rays that are emitted are characteristic of an element's atomic structure to be identified uniquely from one another. Compositional analysis of soda-lime glass and morphological data of the fabricated samples were performed with a scanning electron microscope Zeiss FESEM-ULTRA Plus (fig. 3.8) [Fesem]. Table 3.6 shows the FESEM specifications. Figure 3.7 Zeiss FESEM-ULTRA Plus. Materials, instrumentation and characterization methods CHAPTER 3 37 Table 3.6 FESEM ULTRA plus specifications Specifications FESEM ULTRAPLUS Resolution (optimal WD) 0.8 nm @ 30 kV (STEM mode) 0.8 nm @ 15 kV 1.6 nm @ 1 kV Magnification 12 - 1,000 ,000 x in SE mode / 100 - 1,000 ,000 x with EsB® detector Emitter Thermal field emission type, stability >0.2%/h Acceleration Voltage 0.02 kV - 30kV Probe Current Configuration 1: 4 pA – 20 nA / Configuration 2: 12 pA – 100nA Detectors EsB® detector with filtering grid (0 – 1500 V), High efficiency in-lens SE detector, Chamber mounted Everhart-Thornley detector, Integrated AsB® detector Chamber 330mm (Ø) x 270 mm (h), 3 EDS ports 35° TOA, CCD-camera with IR illumination Vacuum System Complete dry pumping system composed of Backing Pump, Turbomolecular Pump and Ion Getter Pump, Automatically controlled Quiet Mode to switch off Backing Pump Charge Compensator Fully automated and neumatic retractable local gas injector Specimen Stage 5-Axes Motorised Eucentric Stage X = 130 mm, Y = 130mm, Z = 50mm, T = -3 to 70° R = 360° (continuous) 6-Axes Eucentric Stage X = 100 mm, Y = 100mm, Z = 42 mm, Z’ = 13 mm, T = -4 to 70° R = 360° (continuous) Image Processing Resolution: Up to 3072 x2304pixel, Image Display High end 19" flat panel TFT colour display monitor with SEM image displayed at 1024 x 768 pixel System Control SmartSEM®* with Windows®XP. Space Requirement Minimum footprint: 1.97 m x 1.73m, Minimum working area: 3.5 m x 5.0 m 3.5.3 Optical microscope. The diameter and optical images of the marks and elements fabricated by laser ablation of glass materials was performed with an optical microscope Nikon MM400 (fig. 3.8)[Microscope]. This instrument was also used for fast visual analysis of Materials, instrumentation and characterization methods CHAPTER 3 38 the sample, allowing for their selection / rejection before continuing with the thermal treatment. [MM-400]. Table 3.7 shows the specifications. Figure 3.8 Optical microscope Nikon MM-400 Table 3.7 Optical microscope MM-400 specifications. Specifications MM-400/SLU Z-axis movement Manual(dual side coarse/fine focus knob) MM controller backpack interface Built-in Optical head Y-TB binocular eyepiece tube, LV-TI3 trinocular eyepiece tibe, LV-TT2 tilting trinocular eyepiece tube Z-axis linear scale Built-in Eyepiece CFI10x (Field nº 22), CFI10x CM (Field nº 22) Objective CFI60 LU Plan Fluor EPI series, CFI60 LU Plan Fluor BD series, CFI60 LPLAN EPI CR series Stage 6x4, 4x4, 03L, 2x2 Light source Diascopic LED diascopic illuminator (standard), 12V-50W Episcopic White LED illuminator LV-EPI LED, Motorized universal eEPIilluminator LV-U EPI2A, Universal Illuminator with focusing Aid LV-U EPI FA Max. workpiece height 150mm 3.5.4 UV/VIS Spectrophotometer. The optical transmission of glass materials and the optical elements fabricated were analyzed with a UV/VIS spectrophotometer PerkinElmer Lambda 25 covering Materials, instrumentation and characterization methods CHAPTER 3 39 a spectral range between 200 and 1100 (fig 3.9) [Spectrophotometer]. Table 3.8 shows the UV/VIS Perkin Elmer Spectrophotometer specifications. Figure 3.9 UV/VIS Perkin Elmer Spectrophotometer. Table 3.8 UV-VIS-IR Perkin Elmer Spectrophotometer specifications. Specifications Lambda 25 Wavelenght range 190-1100 nm Bandwidth 1 nm Stray light At 220 nm (Nal) < 0.01%T At 340 nm (NaN02) < 0.01%T At 370 nm (NaN02) < 0.01%T At 200nm (KCl) < 0.01%T Wavelength accuracy At D2 Peak ( 656.1nm) ± 0.1nm Wavelength reproducibility 10 measurements at 656.1nm ± 0.05 nm Photometric accuracy At 1 A using NIST 930D filter ± 0.001 A At 2 A using NIST 1930D filter ± 0.005 A Potassium dichromate ± 0.010 A Photometric reproducibility Maximum deviation of 10 measurements at 1 A < 0.001 A Photometric stability Stability at 1 A, at 500nm < 0.00015 A with 2 sec response time A/hour Photometric noise at 500 nm(RMS) Noise 500nm/0 A RMS Slit 1 nm < 0.00005 A Baseline flatness Slit 1 nm ±0.001 A Materials, instrumentation and characterization methods CHAPTER 3 40 3.5.5 Beam profiler The irradiance distribution of the laser beam and the microlens spot size and irradiance at focus were analyzed using a beam profiler BP-109UV, which is a high-precision instrument that can analyze the power distribution of laser beams with diameters from 10 μm to 9 mm. Figure 3.10 shows the beam profiler [Beam Profiler]. Table 3.9 shows the Beam profiler BP-109UV specifications. a) b) c) Figure 3.10 a) BP-104 beam profiler image, b) beam profiler internal mechanism and c) 3D profile of irradiance distribution provided by the beam profiler. Table 3.9 Beam profiler BP-109 UV-IR specifications. Specifications BP-104 UV-IR Wavelength Range (nm) 200 - 1100 Detector Type Si - UV Enhanced Aperture Diameter 4 mm Scan Method Scanning Slits Slit Size 2.5 µm Min Beam Diameter 10 µm Max Beam Diameter 4 mm Scan Rate 1.0 - 20.0 per s (continuously variable) Sampling Resolution 0.5 - 38 µm (depending on scan rate) Materials, instrumentation and characterization methods CHAPTER 3 41 Power Range 10 nW - 10 W (depending on beam diameter and model) Amplifier Bandwidth 10 to 150 kHz in 10 kHz Steps (@ -1 dB) Sample Frequency 0.0625 - 1.0 MHz Dynamic Range 72 dB (Amplifier Switchable) Signal Digitization 16 bit Head Size Ø 80 mm x 60 mm (including rotation mount) Minimum Pulse Rate 10 Hz (300 kHz if using the M² Option) Software Displayed Parameters/Features X-Y-Profile, Centroid Position, Peak Position, Pseudo 3D Profile, Beam Width Clip Level/Second Moment (4σ), Gaussian Fit Applicable, Colored Pass/Fail Test Compliant to Norm ISO 11146 (Beam Widths, Divergence Angle and Beam Propagation Factor) General System Requirements Windows™ 2000/XP/Vista or later, USB2.0 port recommended, 120 MB HD, 512 MB RAM M² Analysis System Compatible M² Options BP1M2-xx Series, M2SET-xxx Series Compliant to Norm ISO 11146 Measured Parameters** M², Waist Width, Waist Position, Rayleigh Length, Divergence, Beam Pointing, Waist Asymmetry, Astigmatism Ablative process of soda-lime glass. CHAPTER 4 48 Figure 4.3 Images of an array of holes machined on soda-lime glass at 1 kHz, scan speed 100 mm/s and different fluences, a) 130 J/cm2, b) 126J/cm2, c) 112J/cm2, d) 109J/cm2 e) 103 J/cm2 and f) 100 J/cm2 As it can be appreciate on figure 4.3, which is in agreement with figure 4.2, each point at figures is the result of the ablation of one pulse shot at the surface of the glass. As the energy per pulse is equal for all the points, the places where no holes appears means that the interaction between laser beam and substrate is difference on points at the surfaces of the glass. Taking into account that we have demonstrated the role of impurities over the ablation process of glass and, assuming that the impurities incorporation is the result of having the glass floating in a Sn bath during the fabrication process, we can relate that the non-uniform distribution of impurities in the glass leads to differences on the ablation threshold and, as presented in figures 4.2 and 4.3 to non-uniform distribution of the holes. a) b) c) d) e) f) g) h) Ablative process of soda-lime glass. CHAPTER 4 49 4.3 Role of surface Roughness. It has been reported through measurements of the polarization of light reflected from particulate surfaces that there is also a component of light reflected that has not been transmitted but has only interacted with the constituent particles [Dollfus1985]. We will proof the following hypothesis: the incoming light can be scattered by microscopic surface roughness and the defects make possible that light interacts with the surface of transparent materials, which leads to an enhanced ablative process. After testing some samples it was realized that the roughness has a very important role in the mechanism of the process. In order to demonstrate it, some experiments were performed over unprocessed glass and a rough glass surface pre-treated in a polishing machine. The polishing process creates a surface with a Roughness Average (Ra) of 554 nm (Fig. 4.4). Figure 4.4 Confocal image of roughness generated at the surface of glass. It is found to consist of craters machined in the glass in the range of diameter 100-400 nm and height 100-600 nm. Ablative process of soda-lime glass. CHAPTER 4 50 Over the pretreated glass, it was fabricated a line at a fluence (15.66 J/cm2) below the ablation threshold of soda-lime glass for non Sn side. Figure 4.5 shows a 3D confocal image and a profile of a line fabricated over the pre-treated glass. The laser used was an IR Nanosecond laser operating at 1064nm with pulse duration of 20 ns. a) b) Figure 4.5 a) Profile and b) Confocal top view image of one line fabricated at surface of structured glass, the blue square in figure a) shows the topography of the roughness generated at glass surface with the polishing machine. The laser parameters were 5 W, 10 kHz, scan speed of 100 mm/s. As it can be appreciate in figure 4.5, it was possible to fabricate a line on the top surface of a pre-treated glass at fluence of 15.66 J/cm2, which is two orders of magnitude below the ablation threshold for soda-lime glass with no Sn impurities. The results obtained show that the roughness of the glass surface is significant for reducing the ablation threshold. Based on this experiment, we fabricate a line over pre-treated surface by applying various laser passes. Figure 4.6 shows the depth as a function of number of passes over a pre-treated glass substrate. Ablative process of soda-lime glass. CHAPTER 4 51 Figure 4.6 Depth as a function of number of passes for roughened glass substrate. From the slope obtained on figure 4.6 it can be observed the evolution with the number of passes. The depth of the mark increases with the number of passes made with the laser, reaching a maximum value of depth 10µm after 18 pass, after this value, an increasing of the laser pass does not becomes in a increasing of the mark depth. That behavior is due to the fact that the ablation plume has not energy enough to take out the material detached on the ablation process. 4.4 Influence of pulse width. The influence of pulse width was analyzed in the range of IR wavelength for femtosecond, picosecond and nanosecond lasers. Figure 4.7 shows the morphology obtained for each pulse width at the fluence threshold. 0 2 4 6 8 10 12 14 16 18 20 22 0 10 Numberof passes 8 6 4 2 Depth (µm) Ablative process of soda-lime glass. CHAPTER 4 52 a) b) c) Figure 4.7 Morphology of the mark created after one laser shot on soda-lime glass for a) nanosecond laser, b) picoseconds laser and c) femtosecond laser. As it can be appreciated in figure 4.7, the morphology of the mark is different for each pulse width, which is expected to be related with the different mechanism involved in the ablation process. Since the heating time is equivalent to the pulse width, different interactions times are involved in transferring the energy from the laser to the lattice of the material. As the laser energy is principally absorbed by the electrons, the electron-phonon/lattice coupling time is in the range of 1 picosecond and the overall relaxation time is in the range of 200 picoseconds. It is expected that when using nanosecond pulse width, the laser energy is absorbed while these interactions take place; it is the accumulative or thermal effects which will lead to ablation. When using a femtosecond pulse width, the interaction time of laser pulses with the materials is less than the electronphonon interaction and relaxation times, so the electrons have time to absorb, transfer and relax before the next laser pulse comes; here the damage may accumulate in the material but the process will be less thermal [Krzystof2005, Shinonaga2009, Rudolph1999, Stuart1996]. In order to get more useful data for a better understanding of the ablative process, investigations and comparison of several glasses (borosilicate, fused 10 µm 10 µm10 µm Ablative process of soda-lime glass. CHAPTER 4 53 silica, sapphire) with the results obtained for soda-lime glass were done. Figure 4.8, shows the ablation threshold data obtained for different glass samples for nanosecond, picoseconds and femtosecond laser pulses. Figure 4.8 Ablation threshold data obtained for different glass samples, for nanosecond, picoseconds and femtosecond laser pulses. In figure 4.8 it can be appreciate that the best result in terms of energy reduction is obtained for nanosecond pulses in the case of soda-lime glass with Sn impurities, which reduces the ablation threshold by one order of magnitude. This enhancement is very important since most of the glasses do not absorb IR wavelengths and most of the micromachining nanosecond IR lasers do not have enough power for machining the glass, or work just in the limits, so the important reduction in terms of energy that offer the ablation through the impurities is of crucial importance. In case of picoseconds, the ablation threshold is one order of magnitude below than the ablation threshold of nanosecond laser for all the glasses, excluding sapphire. Femtosecond pulse width presents the lowest ablation threshold, obtaining a reduction of two orders of magnitude. For 954.86 Borosilicate 923.23 Soda-lime (Side without Sn) 112.68 Soda-lime (Side with Sn) 11.02 Fused silica 9.40 Borosilicate 9.54 Soda-lime (Side without Sn) 8.57 Soda-lime ( Side with Sn) 3.60 Fusedsilica 4.18 Sapphire 2.98 Borosilicate 2.90 Soda-lime (Side withoutSn) 1.98 Soda-lime (Side with Sn) Sapphire Sapphire Fused silica Nanosecond Femtosecond Picosecond 0 1000 J/cm2 100 Ablative process of soda-lime glass. CHAPTER 4 54 picosecond and femtosecond no significant reduction on the ablation threshold was observed for the soda-lime glass with Sn impurities, compared with Sn without impurities. It can be explained by the interaction mechanism of laser pulse and glass and its pulse width dependence. On glass materials, the absorption of light, which leads to the ablation, comes from a photophysical response, which involves non-linear and linear effects. The pulse duration and the material response time are important factors that will determine the kind of mechanism that leads to ablation of material. Whit long pulse (nanosecond) laser processing, the dominant process involved is the heating of the target material through the liquid phase to the vapor phase, resulting in expansion and expulsion of the desired target material. The Sn impurities incorporated in soda-lime glass acts as seed electrons for initiating the avalanche ionization and the ablation process. The duration of the pulses allows for a long avalanche which results in a enhancement mechanism of the ablation process with a reduction of one order of magnitude on the ablation threshold of soda-lime glass. On the other hand, when the laser field strength is very high, the bounded electrons of the transparent material can be directly ionized through multi-photon absorption. It results in a direct vaporization of the target surface. In this case, the soda-lime glass absorbs the incident laser energy preferable due to multiphoton ionization with low contribution of the electron avalanche ionization due to the short irradiation interval. Therefore the impurities in this case do not present any significant improvement, or at least it is not appreciated from our experiments. Ablative process of soda-lime glass. CHAPTER 4 55 4.5 Laser ablation parameters optimization. Taking into account the results obtained in section 4.1 and 4.2, we selected the soda-lime glass as suitable material for fabricating microlens and microchannels with applications on micro-optics and microfluidics at low cost. The optimization of laser parameter comes from analyzing the ablation threshold, the diameter of the mark generated with the laser beam on the soda-lime glass and the pulse overlap, which determine number of pulse per area. In addition, the morphology of the mark generated at surface during laser ablation and the analysis of debris deposition will help us to determine the best laser parameters needed for fabricating different devices. To this purpose, it was investigated, for an indentified threshold fluence and beam diameter, the morphology of damage created at the surface of the glass, using these data for optimizing the designing parameters of the posterior fabrication of micro-optics and microfluidics devices. 4.5.1 Ablation threshold and beam diameter A simple theoretical value for a single-shot threshold fluence is found by assuming that the light is absorbed according to Beers Law and ablation occurs when the energy per unit volume exceeds the energy necessary for the material to undergo heating, melting and evaporation. In terms of energy the ablation threshold was determined using the Eq. (4.1) [Ben-Yakar2004], (4.1) Ablative process of soda-lime glass. CHAPTER 4 56 Where Eth is the energy per pulse at the onset of ablation and is the radius at 1/e2 of the Gaussian laser beam. The beam spot size, which is required for determining the fluence threshold, was determined using the squared diameter (D2) of the ablated areas [BenYakar2004]. The D2 of the ablated areas is plotted as a function of the peak fluence of the laser Fpeak. The slope of the linear fit permit us to know the beam radius at the surface, ω0. (4.2) In order to estimate the fluence, the precise knowledge of ω0 is required. Since it is easier to measure the energy per pulse, E pulse, experimentally, we replace the fluence ratio in Eq. (4.3) by the ratio of the pulse energies, yielding: (4.3) Because the measured Gaussian spot size on the surface is not necessarily equal to the calculated spot size at the focal point of the objective, D2 calculation is more realistic than theoretical calculations of beam spot sizes. Figure 4.9 shows this linear relation between the squared diameter and the natural logarithm of the average laser fluence in our measurements for the nanosecond laser. A value of 21.65 µm was obtained. Ablative process of soda-lime glass. CHAPTER 4 57 Figure 4.9 Linear relations between the squared diameter and the natural logarithm of the average laser fluence in our measurements for nanosecond. 4.5.2 Mark analysis and pulse overlap. For determining the size of the mark made by the laser beam in the glass substrate, a crater was made in the glass after firing one shot at a fluence of 112 J/cm2. The size of the crater was analyzed with a confocal microscope obtaining 15 µm width and 8 µm height for soda-lime glass. Figure 4.10 shows the morphological data, height and diameter, of the mark obtained after one laser shot on soda-lime glass. These data also provide the useful information for calculating the pulses overlap, and so the number of pulses per area. a) b) Figure 4.10 a) Profile and b) confocal image of a crater made on glass substrate by one shot of the laser beam. 10 100 0 500 1000 1500 2000 2500 3000 D2 ( µm) ln (F peak / Fth ) j/cm2 5 Ablative process of soda-lime glass. CHAPTER 4 64 4.6 Enhancement of the ablative process using a thin absorber layer. Considering laser-matter interaction, an attempt is made to identify the mechanism that leads to the ablation of the transparent materials using a thin film coating at different pulse durations, so three factors affecting the ablation region were identified on literature and considered in this work. The first one is related with structural defects in the aluminum thin film (roughness) which enhances absorption in dielectric. The second one is related with impurity incorporation which increases the absorption of laser energy, during later part of the pulse or for subsequent pulse. The last one is related to the charge transfer from the thin film into dielectric which leads to avalanche absorption mechanism. The use of a solid thin film as an absorber layer of the laser irradiation was used in this work to aid the ablation process. To this purpose a 15 nm thin aluminum layer was deposited over different glass materials using a physical vapour deposition machine (Balzers BAE 250 coating system). All of these factors are related with the coupling of light at the glass surface and the transmission of energy from the laser to the aluminum that interacts with the glass. Some experiments were developed for trying to get the first approach to our hypothesis. Figure 4.16 shows a diagram of the process involved in these hypotheses. a) b) c) Figure 4.16 Sketch of the mechanism that leads to the laser ablation of glass using the aluminum thin film, a) structural defects (roughness), b) impurity incorporation and c) charge transfer Glass Substrate Plasma Plume Laser Pulse,IR Thin film Plasma Plume Laser Pulse,IR Metal ions Glass Substrate Thin film Laser Pulse,IR Plasma Plume Thin film Glass Substrate Laser Pulse, IR Shockwaves Ablative process of soda-lime glass. CHAPTER 4 65 Figure 4.17, shows the ablation threshold data obtained for coated and uncoated different glass samples, for nanosecond, picoseconds and femtosecond laser pulses. 954.86 Borosilicate 923.23 Soda-lime (Side without Sn) 112.68 Soda-lime (Side with Sn) 11.02 Fused silica 9.40 Borosilicate 9.54 Soda-lime (Side without Sn) 8.57 Soda-lime ( Side with Sn) 3.60 Fusedsilica 4.18 Sapphire 2.98 Borosilicate 2.90 Soda-lime (Side withoutSn) 1.98 Soda-lime (Side with Sn) Sapphire Sapphire Fused silica Nanosecond Femtosecond Picosecond 0 1000 J/cm2100 33.62 Fused silica 27.63 Sapphire 9.34 Borosilicate 7.32 Soda-lime (Side without Sn) 6.96 Soda-lime (Side with Sn) 9.96 Fused silica 9.58 Sapphire 10.53 Borosilicate 9.46 Soda-lime (Side without Sn) 7.05 Soda-lime (Side with Sn) 5.12 Fused silica 6.34 Sapphire 8.12 Borosilicate 4.21 Soda-lime (Side withoutSn) 4.52 Soda-lime (Side with Sn) Figure 4.17 Ablation thresholds for different glass materials at different pulse duration. Black text relates the ablation threshold for unprocessed glass and blue text refers to the ablation threshold for glass covered with the absorber layer. It is clear that the best result is obtained for nanosecond pulses. In this case the ablation threshold is reduced by two orders of magnitude for borosilicate, fused silica and sapphire and one order of magnitude for soda-lime glass. In case of picosecond pulses it is slightly reduced in some glasses, less in borosilicate in which the overall value for ablation threshold is a little bit higher. For femtosecond pulses there is no significant reduction in the ablation threshold and the enhancement is minor. The best enhancement in terms of energy reduction was obtained for nanosecond laser and the glass covered with the aluminum layer, presenting low ablation threshold, which is very important since most of the glasses do not absorb IR wavelengths well and so, in order to ablate the glass, it is necessary to use high Ablative process of soda-lime glass. CHAPTER 4 66 energy levels, which is provided by femtosecond and picoseconds lasers. Besides, most of the micromachining nanosecond IR lasers do not have enough power for machining the glass, or work just in the limits, so the important reduction in terms of energy that offer this technique is very useful for machining the glass. Although, the picosecond and femtosecond pulses do not present important results regarding energy reduction, the data obtained is very useful for developing the understanding of the ablation process developed in this work, which will be analyze in next section. 4.6.1 Factors affecting ablation region: Role of roughness The incoming light can be scattered by microscopic surface roughness and defects, this makes it possible for the light to interact at the surface of transparent materials which leads to an enhanced ablative process. After testing some samples we can assert that the roughness has a very important role in the mechanism of the process. In order to probe that, a line was fabricated over an absorptive aluminum layer and a glass surface pre-treated in a polishing machine for generating a rough surface. The roughness average (Ra) of aluminum layer was estimated to be 30 nm, the roughness average of pre-treated glass surface, 600 nm and the Ra of soda-lime unprocessed glass was 2.1 nm. The objective is to relate the size of the particles of aluminum layer with the size of the particles created in the glass using the polishing process and so related both with roughness. The size of aluminum nanoparticles was determined with a scanning electron microscope, obtaining values diameter around 20 nm for aluminum. The polishing treatment generates structures at surface of glass in the range of diameter 100-300nm and height 100-600 nm. The morphological data is shown in figure 4.18 and 4.19 respectively. Ablative process of soda-lime glass. CHAPTER 4 67 Figure 4.18 SEM image of aluminum layer; it is found to consist of an aggregate of nanoparticles for the 15nm – 40nm thickness range used in this study. Figure 4.19 Confocal image of roughness generated at the glass surface. It is found to consist of craters from polishing of the glass in the range of diameter 100-300 nm and height 100-600 nm. Ablative process of soda-lime glass. CHAPTER 4 68 The line was fabricated below the ablation threshold for uncoated glass 15.66 J/cm2, the results obtained demonstrate that both the aluminum layer and the roughened glass surface are significant for reducing the ablation threshold. In order to get more data, the depth of the line as a function of number of passes is plotted in figure 4.20 Figure 4.20 Depth as a function of number of passes for (red line) glass covered with aluminum layer and (green line) roughened glass substrate using IR laser parameters, pulse width of 20ns, repetition rate 1 kHz and fluence 15.66 J/cm2. The data suggests that the evolution with the number of pulses saturates after a specific depth (10 µm for aluminum layer and 8.5 µm for pre-treated glass). This saturation on depth is related with the defocusing of laser beam over the substrate by increasing the depth of the channel, leading to less energy per area, which means that there is no enough energy to take out the material detached on the ablation process. So in summary the first factor try to relate the roughness at the surface as a scattering centre that improves the coupling of light into the material enabling the subsequent transfer of energy from the light to the substrate. Light reflected from 0 2 4 6 8 10 12 14 16 18 20 22 0 10 Depth (µm) Number of passes Structured glass Aluminumlayer 2 4 6 8 Ablative process of soda-lime glass. CHAPTER 4 69 a smooth metallic surface is confined to a specularly reflected beam. The intensity of this reflected beam is given by Fresnel's equations and it depends on the polarization state of the incident light, the angle of incidence, and the optical properties of the metal. Typically this is an ideal situation as for thin metals deposited on glass substrates the coatings are not smooth. So when light is reflected from rough surfaces, two new phenomena occur. Firstly, the specular reflection does not occur; the reflected scattered light is directed away from the direction defined by the specularly reflected beam resulting in a reduction in the specularly reflected intensity. Secondly, additional excitations or currents may be generated across the pits and bumps on the roughened surface. These excitations, known as surface plasmons, can give rise to additional fields which mix with the specular and scattered intensities forming surface plasmon polaritons. The study of the surface plasmon polaritons are outside the scope of this study, but it is proposed that the scope for generating plasmonic excitations at the interface of a dielectric material with a random array of aluminium particles is likely to form part of the enhanced mechanism [Villesen2012]. As yet there is no firm evidence to support this hypothesis, the following points are made however to justify further investigations:  The enhancement mechanism only works for thin metal coatings consisting of a random arrangement of nanoparticles.  The nanoparticle sizes are smaller than the wavelength of light indicating that plasmonic behavior is likely to occur. Ablative process of soda-lime glass. CHAPTER 4 70  The coatings have a reduced reflectivity that evidences possible plasmonic absorption.  The optical penetration depth of the thin aluminium layer indicates that there is scope for the incident electromagnetic field and the laser to interact at the metal dielectric interface.  The surface plasmon polariton mode decays non-radiatively although the different decay routes have not been investigated. 4.6.2 Factors affecting ablation region: Role of diffusion The diffusion of metal ions into the glass is of relevance to the manufacture of glass-based optical materials which is expected to maintain the same properties before and after laser machining. Because of this, the structural changes of the glass during the covering process and after laser exposure have been analyzed, in order to know if the aluminum diffuses into the glass. In this case the diffusion will contaminate the sample yielding to a reduction of our technique capabilities. 4.6.2.1 Diffusion at covering the glass To investigate the diffusion of metal ions into the glass during the process of covering the glass with a thin aluminum layer, the glass is covered with an aluminum layer of 15-20 nm thickness, which is then removed using a mixture of acids, composed by nitric acid and hydrochloric acid, in a volume ratio of 1:3. Then the etched surface is analyzed with the EDX technique. Table 4.2 shows the composition of soda-lime glass before and after the treatment explained above. Ablative process of soda-lime glass. CHAPTER 4 71 Table 4.2 Composition of soda-lime glass before and after the treatment. Before treatment After treatment Element Weight% Element Weight% O 50.58 O 50.25 Na 7.22 Na 8.08 Mg 2.27 Mg 2.19 Al 0.47 Al 0.49 Si 30.70 Si 32.08 Ca 4.44 Ca 4.87 Since the variations of aluminum percentage measured at different points of the glass previous to deposition of aluminum layer was in the order of 0.03 %. From table 4.2, it can be appreciated no significant additional aluminum is observed in the glass, so we can say that the coating process do not contaminate the glass. 4.6.2.2 Diffusion during ablation. Next the diffusion of aluminum ions into the glass during the ablation of the glass was investigated. In order to get information about the diffusion of metal ions into the glass, a channel was fabricated over the glass using the aluminum layer; after this the aluminum layer was etched and then analyzed using EDX. Figure 4.21 shows a top view of the soda-lime glass and the points selected for obtaining the composition data using EDX. Spectrums 1, 2 and 3 were obtained in the bottom of the channel and spectrum 4, 5 6 and 7 were obtained out of channels and over the surface of unprocessed glass. Ablative process of soda-lime glass. CHAPTER 4 72 Figure 4.23 SEM picture of soda-lime glass and the points selected for analyzing the variations in composition after laser ablation. Table 4.3 shows the relative compositions obtained in different places of the laser etched glass, locations 1,2, and 3 shows the EDX analysis at the bottom of the channels where laser –aluminum interactions occurred , locations 4,5,6,7 shows EDX analysis outside of the channels where no laser-aluminum interaction occurs. Table 4.3 Spectrums obtained for different places of the glass substrate using EDX. Element Spectrum 1 Spectrum 2 Spectrum 3 Spectrum 4 Spectrum 5 Spectrum 6 Spectrum 7 Glass O 49.22 50.54 50.04 49.69 49.35 49.22 50.21 50.58 Na 7.52 8.46 7.05 8.12 7.55 7.52 8.02 7.22 Mg 2.34 2.53 2.25 2.03 2.14 2.34 2.38 2.27 Al 0.41 0.54 0.52 0.46 0.59 0.41 0.45 0.47 Si 30.66 32.13 30.35 28.27 28.22 30.66 30.60 30.70 Ca 4.72 4.75 4.34 3.98 3.94 4.72 4.71 4.44 Ablative process of soda-lime glass. CHAPTER 4 73 Comparing those results, and within the sensitivity of EDX technique, we can conclude that “little or no diffusion” occurred in the laser treated glass. This initial result is in agreement with a simple model for diffusion. The diffusion of the metal ions in a glass matrix is governed by Fick’s second law. In one dimension into the depth of the material, the profile metallic impurities are given by equation (4.5). 2 2 x N D t N      (4.5) where N is the concentration of the metal ions, D is the diffusion coefficient, t is the time variable and x represent the depth magnitude. The solution of this differential equation requires consideration of the appropriate boundary and initial conditions. The solution provides an estimation of the impurity distribution, in depth and time, for the metal ions in the glass material. We consider the initial case of a constant surface concentration, where a metal layer deposited on the glass acts like an unlimited source of metal dopants, providing metallic ions into the glass materials. In this case the solution is given by the complementary error function Eq.(4.7)[Sze2001] where, (4.7) Equation (4.7) describes the distribution of metal ions in the glass material with depth and time following exposure of the thin film / glass matrix to an initial laser pulse. N0 is the initial metal ion concentration at the surface. N0 can be either represented by the solid solubility of aluminum in the glass material or by the Microstructuring of glass: fabrication and characterization CHAPTER 5 80 5.1 Introduction. A direct-write technique for fabricating plano-convex microlenses and microchannels based on the ablation of a soda-lime glass substrate with a laser beam of circular Gaussian profile is presented. The laser setup consists of a nanosecond Q-Switch Nd:YVO4 laser operating at the fundamental wavelength of 1064 nm combined with a galvanometer system. Laser parameters selected as suitable from results obtained in chapter 4, were an average power of 7 W and a repetition rate of 10 kHz. A wet chemical etching technique (HF) was used successfully to carry out smoothing and cleaning of surfaces fabricated on the soda-lime glass. Following cleaning processing a thermal treatment in an oven was used for reflowing the glass surface. By applying the thermal treatment a modification of the shape of the elements occurs, as well as an improvement of the morphology in terms of roughness and elimination of micro crack defects generated during laser ablation. The thermal treatment was applied using a Heraeus Mufla furnace, equipped with a thermal ramp that allows us to control the heating speed of the glass. 5.2 Laser microstructuring of glass The laser setup consists of a nanosecond Q-Switch Nd:YVO4 laser operating at a wavelength of 1064 nm combined with a galvanometer system for addressing the output laser beam. The laser is focalized onto the sample with a flat-field lens, with effective focal length 100 mm, that provides a uniform irradiance distribution on glass substrate over an area of 80x80mm2 (Figure 5.1). Microstructuring of glass: fabrication and characterization CHAPTER 5 81 Figure 5.1. Laser set-up for microstrutring glass. In order to design and fabricate microlenses and microchannels a previous study was made by focusing one laser shot on the glass substrate [section 4.5.2 of chapter 4]. The mark obtained was analyzed with a confocal microscope SENSOFAR 2300 Plµ. The initial characterization was used to determine the appropriate laser parameters needed for the fabrication of the microlens arrays and microchannels on the soda lime glass substrate. For fabricating microlens, each cylindrical structure that will lead to microlens after a thermal treatment is obtained by the ablation of a circular trench formed by moving the laser beam, using the galvanometer mirror system, relatively to a stationary glass sample (Figure 5.2c). A time of 10 seconds is required to expose an area of 2x2 mm2. Figure 5.2a and 5.2b show the scheme of the hexagonal packed array designed for fabricating the microlens array. Microstructuring of glass: fabrication and characterization CHAPTER 5 82 Figure 5.2 a) Hexagonal close packed pattern, b) Microlens array arrangements: diameter and pitch. c) Lateral view of the cylindrical trench formation. In the case of microchannels, each structure is obtained by the ablation of individual lines on different configurations on soda-lime glass substrate using a galvanometer system. Figure 5.3 shows the cylindrical structures that will lead to microlens and the microchannels obtained by laser direct write. The samples were characterized by a confocal microscope SENSOFAR 2300 PLµ. a) b) Figure 5.3. 3D confocal image of a) cylincrical post b) microchannel fabricated on soda-line glass by laser–direct write. The first experimental approach for obtaining uniform channels and cylindrical structures was to check different scan speed for getting the optimal pulses Galvanometer Initial microlens Nd:YAG 4laser Lens 60 µm 60 µm Beam motion 50 µm 10 µm 50 Y (µm) X (µm) Z (µm) 6 3 0 -3 40 70 100 130 80 110 170 140 50 Y (µm) X (µm) Z (µm) a) b) Figure 5.2 (a) Confocal image of the cylindrical posts obtained by laser ablation and b) after 2 hours into a oven at 670ºC. It can be observed in Figure 5.2 that the ablation creates cylindrical posts with a flat top surface without any refracting power. The conversion of the initial pattern of glass posts fabricated by laser ablation into the appropriate shape for the microlenses starts at temperatures above the glass transition temperature, Tg, where the viscosity of the glass is reduced and thus the surface tension of the melted material induces the modification of the surface shape through the redistribution of the material from the top sides of the trench to the bottom of the crater. 125 110 95 80 X (µm) 60 75 45 30 Y (µm) 65 -3 -8 Z (µm) Microstructuring of glass: fabrication and characterization CHAPTER 5 83 overlapping and laser parameters. Section 4.5.2 of chapter 4 describes the previous experiments performed for obtaining the optimal laser parameters. After laser processing, in order to obtaining microlens and microchannels, a chemical etching and a thermal treatment in an oven was applied to the elements to reflow the glass surface and improve the topography in terms of roughness. The chemical etching and thermal treatment mechanism are studied in the next sections. 5.3 Thermal treatment. The cylindrical structures and microchannels fabricated by direct laser write were subjected to thermal treatment into a mufla oven Hearaus and a Vita Vacumat 6000M (for specifications see section 3.3 of chapter 3). Regarding the mechanism of thermal reflow for microlens fabrication, the displacement of material from the upper surface and the consequential accumulation at the bottom surface of the feature leads to a reduction in height of the island structure generated by the laser. The approximate diameter is maintained because of the confluence of the material reflowed from the top of adjacent microlens to the bottom of the crater. This effect permits to fill the irregular structure of surface leading to an improvement of the optical and morphological qualities of the glass structures. Figure 5.4 shows the mechanisms involved in the process, which leads to morphological shape modification from the initial cylindrical post to a microlens formation. The black line shows the profile of the initial post obtained by laser ablation and dot lines the profile of resulting microlens after thermal treatment. Microstructuring of glass: fabrication and characterization CHAPTER 5 84 Figure 5.4 Material displacement mechanisms for fabricating microlenses. It can be observed in Figure 5.4 that the laser ablation creates cylindrical posts with a flat top surface without any refracting power. The profile of the initial laser etched microlens is similar to a cylindrical shape with the corners of the features being partially rounded. The conversion of the initial pattern of glass posts fabricated by laser ablation into the appropriate shape for the microlens starts above the glass transitions temperature (Tg =564oC), where the viscosity of the glass is reduced and thus the surface tension of the melted material induces the modification of the surfaces shape through the redistribution of the material from the top sides of the trench to the bottom of the crater. Based on this, it was decided to apply the thermal treatment for temperatures above Tg, (620ºC, 630ºC, 640ºC, 650ºC, 660ºC, 670ºC, 680ºC) during 2 hours. The morphology in terms of shape was analyzed with confocal microscope SENSOFAR 3200 Plµ. For a temperature of 620ºC or lower, there was almost no change in the microlens shape. But at temperatures higher than 670ºC, the initial cylindrical surface profile becomes increasingly flattened, so microlenses were not obtained. Figure 5.5 shows the cylindrical posts obtained by laser ablation and the same element after the oven treatment at 670ºC for 2 hours. The material changes the position from the top to the bottom Before thermal treatment After thermal treatment µm Microstructuring of glass: fabrication and characterization CHAPTER 5 85 Figure 5.5 a) Confocal image of the cylindrical posts obtained by laser ablation and b) after 2 hours of thermal treatment at 670ºC. The material displacement mechanism for microchannel fabrications is similar to that for fabricating microlens. Figure 5.6 shows the material displacement mechanisms for fabricating microchannels. The black line shows the profile of the initial channel obtained by laser ablation and dot lines the profile of resulting microchannels after thermal treatment. Figure 5.6 Material displacement mechanisms for fabricating microchannels. Figure 5.7a and 5.7b shows a microchannel after thermal treatment for 2 hours at 620oC and 670oC, respectively. 0 60 120 180 240 12 8 4 0 -4 -8 X (µm) Beforethermaltreatment After thermaltreatment Thematerial changes the positions from the top to the bottom 1.0 2.0 3.0 4.0 5.0 Microstructuring of glass: fabrication and characterization CHAPTER 5 86 a) b) Figure 5.7 Confocal images of the microchannels when into the oven at a) 620oC and b) 670oC. Since the purpose of thermal treatment in the case of microchannels is to reduce the roughness generated during laser ablation, it is important to find a compromise between shape modification and roughness reduction. Ideally the shape should be maintained while the thermal reflow should reduce the surface roughness. By applying the thermal treatment at 620ºC (during 2 hours) it was able to obtain high quality microchannels maintaining the initial shape but reducing the roughness. At 670ºC the roughness obtaining is similar to the unprocessed glass but the shape changes considerably. In terms of roughness, the thermal treatment permits us to obtain values of roughness in the range of unprocessed glass. Table 5.1 shows the Ra of microchannels obtained at different temperatures. Table 5.1 Roughness average determined bottom of the channels Ra Glass 3.6 nm Ra Glass after laser 2480 nm Ra 620ºC 125.1 nm Ra 670ºC 7.3 nm 120 120 150150 180 210 210 240240 0.2 -0.1 -0.4 -0.7 0.0 -1.0 -2.0 -3.0 Z(µm) 180 Z(µm) X(µm) X(µm) Y(µm) Y(µm) Microstructuring of glass: fabrication and characterization CHAPTER 5 87 5.4 Chemical Etching. In order to reduce the rough surface produced directly through laser ablation, a wet chemical etching technique (HF) was performed and analyzed for microlens diameters of 20 m, 40 m, 60 m and 140 m. The mechanism to modify the surface roughness and the shape depends on the concentration of the acid, on the etching time and on the temperature of the process. Hydrofluoric acid is an etchant which attacks glasses at significant high etch rate. However, HF is not only a strong corrosive, but also highly toxic towards higher concentrations, so the etching process was performed at 10% HF concentration, which in terms of security, means that there is not dangerous percentages of toxic vapor that contaminate the work space. Besides, HF etching was performed with at least one other person in the clean room and was handled in a laminar flow bench, using two pairs of nitrile gloves and eye protection. The purpose of using HF etching was to eliminate the debris deposited over the glass after laser ablation and the secondary post unablated (see figure 5.8) due to the nature process and design configuration. Figure 5.8 shows the results of exposing the initial structures obtained by laser ablation to a 10% HF etching during 5, 10 and 15 minutes. Figure 5.8a shows a 3D confocal image of one microlens before HF etching. Figure 5.8b shows the microlens arrays after HF etching for 5 minutes, 5.8c) for 10 minutes and 5.8d) for 15 minutes. Debris can be observed at the top of the microlens deposited during the laser ablation, secondary post unablated due to the hexagonal packagement of microlens arrays are also observed. Microstructuring of glass: fabrication and characterization CHAPTER 5 88 a) b) c) d) Figure 5.8 3D Confocal image of one microlens a) before HF etching and with, b) HF etching during 5 minutes, c) 10 minutes and d) 15 minutes. As it can be appreciated in figure 5.8, the chemical etching eliminates the debris deposited at surfaces of microlens during laser ablation and the secondary post. A FESEM ultra plus and a confocal microscope SENSOFAR 2300 PLµ were used for depth measurement and 3D mapping of the structures. Figure 5.9 shows a SEM top view image of one microlens before HF etching. Debris deposited at the top of the microlens during the laser ablation can be observed. In figure 5.10 we can see that after chemical etching, they were successful eliminated. 180 0 60 Y (µm) X (µm) 120 180 240 120 60 X (µm) Z (µm) 160 220 280 340 Y (µm) 340 280 220 160 400 460 X (µm) -20 4 - 4 -12 Z (µm) 5 min Secondary post 340 160 220 Y (µm) X (µm) 340 400 460 280 280 220 Z (µm) 160 40 100 160 220 Y (µm) X (µm) 480 420 360 300 540 600 Z (µm) 10 min 15 min Microstructuring of glass: fabrication and characterization CHAPTER 5 89 Figure 5.9 SEM image of the microlens top surface before chemical etching Figure 5.10 SEM image of the microlens top surface after 10 minutes in 10% HF aqueous solution. The elimination of debris will improve the optical quality of the microlens, while if maintained at the top of microlens they would be mixed with the glass material during the thermal treatment resulting in bad optical quality. Figure 5.11 shows a wider top view SEM image of the microlens arrays before HF etching where it is possible to see the secondary posts surrounding the microlens, which are related with the hexagonal packagement of the microlens arrays. Debris Microstructuring of glass: fabrication and characterization CHAPTER 5 96 It can be appreciate in figure 5.18b, the chemical etching makes the initial structures more defined by eliminating bad structures generated and the secondary post generated by the laser ablation process. The secondary post are bigger at diameter 140 µm and interact with material of cylindrical post during material displacement of thermal treatment, so without using the chemical etching it was not possible to obtaining microlens (figure 5.18c). So for obtaining microlens of diameter 140 µm we need to use a combination of laser ablation, chemical etching and thermal treatment. For diameter of 60 µm and 140µm it were obtained focal length values of 1.180 mm +/- 40 µm and 5.100 mm +/- 200 µm, respectively. 5.5 Morphological and optical characterization. In this section we present the morphological and optical characterization of microchannels and microlens. It can be appreciate how the shape of microlens is modified with the thermal treatment and the consequent improvement of the optical properties. It is also presented the evolution of roughness with the thermal treatment. 5.5.1 Microlens. In this section we present the fabrication of arrays of microlenses of diameter 40 m in a lattice with hexagonal symmetry. We present only the characterization of these microlens because in their fabrication is not necessary the use of the chemical etching process. Figure 5.19 presents 3D images of microlens obtained after the thermal treatment at different temperatures. The images were obtained with a SENSOFAR PLµ 2300 confocal. The change in the shape of the microlenses, as well as the filling of the holes between consecutive microlenses is observed. Microstructuring of glass: fabrication and characterization CHAPTER 5 97 Figure 5.19 3D confocal images of microlenses (Diameter: 40 m) obtained for different temperatures of the thermal process. Figure 5.20 presents, for comparison, the profile of one microlens randomly chosen from one of the arrays fabricated at different temperatures. The shape of the microlens evolves from a flat-top post to a quasi-parabolic profile as the Microstructuring of glass: fabrication and characterization CHAPTER 5 98 temperature treatment increases. In the case of 620ºC some irregularities in the surface generated by the laser ablation step were observed. For temperatures of 630ºC and higher these irregularities disappear. The diameter of the microlenses is maintained almost constant over the range of temperatures tested, while the sag decreases due to the redistribution of the material from the top sides of the trench to the bottom of the crater. Temperatures out of this range were also analyzed; it was found that for values lower than 620ºC there were almost no significant changes in the surface shape. By contrast, for temperatures higher than 670ºC and 2 hours of thermal treatment, the initial array of cylindrical surface profile becomes flat, so the microlenses obtained by laser direct-write disappear. Figure 5.20 Cross-sectional profile of the microlens at different thermal reflow temperature. Moreover, the initial irregular structure observed in Figure 5.20 for 620ºC is improved by increasing the temperature, leading to a hexagonal shape for temperatures of 660ºC and 670ºC. This hexagonal profile was generated by enabling the material to reflow, which is related with the initial hexagonal packaging of microlens arrays. The refilling of the interstitial space also contributes to the reduction of the stray-light presented at the focal plane of the microlenses (see Figure 5.21 for details). Microstructuring of glass: fabrication and characterization CHAPTER 5 99 Figure 5.21 Pattern image of the microlens arrays at different temperatures. Images are generated by an optical microscope. The properties of the fabricated microlenses were characterized in terms of its focal length and spot size. The experimental setup for measuring the focal length of the microlenses consisted of a He-Ne laser (632.8 nm), a 40x objective lens and a camera (fig.5.22). Figure 5.22 Experimental setup for measuring the focal length of the microlenses. Microstructuring of glass: fabrication and characterization CHAPTER 5 100 The microscope objective was focused on the surface of the microlenses, this location is recorded, and then the microlenses array is moved along z axis until the image of an arbitrary object is observed. The image is located at the focal plane of the microlenses. The difference between these positions is a measure of the focal distance of the tested microlens arrays. The accuracy of the data acquisition for this setup is ±5 µm. For determining the quality of the focus we used the beam profiler BP109-VIS of Thorlabs. The focus spot size is determined calculating the width at 1/e2. The geometrical - morphological characterization of the microlenses was performed with the confocal microscope. The topographical data provided was used to measure their diameter, roughness and sag, defined as the distance from the top of the microlens to the bottom of the interstitial channel between two adjacent microlens. In table 5.2 we present de results obtained for all this parameters. Table 5.2 Optical and geometrical parameters of the microlenses fabricated at different reflowing temperatures. Reflowing Temperature 620ºC 630ºC 640ºC 650ºC 660ºC 670ºC Focal length (µm) 570±10 510±10 450±10 430±10 330±10 510±10 Focal Spot size (µm) 21.71±1.50 17.14±0.98 8.10±0.15 2.83±0.55 2.82±0.02 2.82±0.02 Diameter (µm) 42.08±4.91 45.77±3.49 46.15±3.39 46.63±3.78 46.48±3.05 47.89±6.65 Sag(µm) 6.58 ±0.57 5.69±0.18 5.52±0.41 4.39±0.22 3.41±0.28 1.32±0.11 Interstitial Roughness(nm) 3000.0±82.0 2950.0±83.0 1540.0±64.0 486.6±13.6 193.0±23.2 54.9±6.0 Lens Roughness(µm) 7.60±0.93 7.18±0.18 7.11±0.46 5.81±0.75 2.30±0.15 2.18±0.34 Microstructuring of glass: fabrication and characterization CHAPTER 5 101 We observe in table 5.2 that the diameter is maintained almost constant over the range of temperatures tested, while the sag, decreases as temperature increases due to the displacement of the melted material. Regarding the interstitial roughness it was observed a significant reduction as the temperature increases. These two last parameters (sag and rougness) are related and respond to the material displacement from the top to the base of the microlenses during the reflowing treatment. While changes in microlenses sag modifie the focusing capabilities, the surface mellowing contributes to the reduction of the stray-light typical of microlens arrays fabricated with laser direct-writing techniques. Table 5.2 also shows a decrease of the focal length of the microlens from 620ºC to 660ºC. This behaviour is related with the reduction of the flatness of the microlens top. At 670ºC it was observed a significant increase in the focal length with respect to that obtained at 660ºC because of the reduction of the spherical aberration of the microlens obtained at 670ºC. In relation to the spot size we observe a decrease as the temperature increases, what confirms the improvement in the microlens focusing power and optical quality. Figure 5.23 shows different images of the focal plane of the microlens arrays reflowed at different temperatures which allows for a visual evaluation of the focusing properties of the fabricated microlens arrays. We observe how the focal spot sizes become significantly smaller and there is a better distribution of irradiance as the reflow temperature increases, providing a significant improvement in their focusing capability. It was also observed how the stray-light reduces as the reflowing temperature increases due to the roughness surface mellowing (see table 5.2), which leads to well-defined spots. Besides, the foci irradiance distribution of the microlens array becomes more regular. Microstructuring of glass: fabrication and characterization CHAPTER 5 102 Figure 5.23 Images at focal plane of the microlens array after thermal treatment at different temperatures. Figure 5.24 shows the variation on the focal irradiance distribution measured with the Thorlabs e-beam profiler, for the microlenses fabricated at different temperatures. This picture gives a visual perspective of the spot size and an approximate impression of the optical aberrations of the microlenses. Comparing the spots it is observed that they become narrower as the temperature increases, resembling the diffraction limited spot in case of 670ºC. Microstructuring of glass: fabrication and characterization CHAPTER 5 103 Figure 5.24 Measured intensity distributions at focal plane for microlens at different temperatures 5.5.2 Optical aberrations: In this subsection it is going to be determined the optical aberrations of the microlens arrays fabricated at 650ºC, 660ºC and 670ºC. We choose these elements because those obtained between 620ºC-640ºC were not significantly influenced by the reflowing process and present poor focusing capabilities. 3D Profiles of the microlens array were obtained with a confocal microscope SENSOFAR 2300 Plµ. The data were used not only for visualization purposes but also for analyzing the optical aberrations of the microlens array by calculating the modal coefficients of the expansion of the height data in terms of the orthogonal Microstructuring of glass: fabrication and characterization CHAPTER 5 104 Zernike polynomial basis. The first step for obtaining the Zernike coefficients is the identification of the pixels that belongs to the different microlenses in the 2D image. The procedure is as follows. First it was built the 2D confocal image. Then it was selected one of the microlenses and correlated with the microlens array image. The resulting image was thresholded in order to select those pixels with a high degree of correlation (which are close to the microlens centre) and then dilated to include most of the pixels that belongs to the microlens. Next we computed the centroid of these pixels in order to locate the centre of the microlens. Figure 5.25 represents the steps described above from the selection of the template to the identification of the microlenses. Figure 5.25 Image processing steps for microlens identification: a) template, b) actual microlens arrays, c) labeled masks and d) identified microlenses. * indicates image correlation. After identifying the microlens it was measured their radius. To do that it was computed the gradient of the 2D microlens array image. By doing this it was enhanced the microlens border (Fig. 5.26). The radius was measured as the distance from the centre to the microlens border. This value was used to define the microlens pupil, inside which the wavefront analysis would be performed. (a) (b) (c) (d) Microstructuring of glass: fabrication and characterization CHAPTER 5 105 Figure 5.26 Border enhanced microlens array image. The surface data inside each microlens pupil was expanded in a series of Zernike polynomials (defined according to the standard OSA-VSIA). We used this base of polynomials because it is orthogonal and some of the terms can be directly identified with common optical aberrations (Z1 and Z2 are prismatic terms, Z3 and Z5 are primary astigmatic terms, Z4 is defocus, Z7 and Z8 primary comatic terms, and Z12 primary spherical (see refs. Malacara2007 and Thibos2002 for the identification of more Zernike terms). The modal coefficients were obtained through the common modal least squares estimation approach (see Eq. 5.1). ˆ -1 S Za b = Z'Z Z' S S = Zb (5.1) where a and b are the column vector of actual and estimated modal coefficients, Z is a MxNs matrix (being M the number estimated modes and Ns the number of pixels inside the microlens pupil, with Ns > 14000 points in all cases), S is the height data and Ŝ is the estimated height. Once we obtain the modal coefficients that allow us to express the surface data in terms of Zernike polynomials we are able to compute different magnitudes Microstructuring of glass: fabrication and characterization CHAPTER 5 112 Figure 5.31 3D confocal image of a microchannel fabricated by direct laser write. The depth and the surface roughness of the etched glass substrates were measured using a Confocal microscope SENSOFAR 2300 PLµ. The resulting channel exhibit a roughness of 640 nm, diameter of 8 µm and height of 3 µm. After laser direct write a thermal treatment in Heraeus mufla oven was applied at different heating times. Figure 5.32 shows the evolution of the topographic profile of the fabricated microchannels at different temperaturess. Figure 5.32 Cross-sectional profile of the microchannels obtained at different thermal reflow temperature (620 ªC, 630 ºC, 640 ºC , 650 ºC, 660 ºC and 670 ºC). It can be appreciated in figure 5.32 that for temperatures of 620ºC or lower, there was almost no change in the surface shape and for reflowing temperature higher 60 75 90 105 120 45 30 Y (µm) -2.0 75 60 -1.0 0.0 -3.0 Z (µm) X (µm) 10 20 30 40 50 60 70 80 670ºC 660ºC 650ºC 640ºC 630ºC 620ºC 0.2 0.0 -0.4 -0.6 -0.8 -1.0 -1.2 -1.4 -1.6 -1.8 -2.0 -2.2 -2.4 -0.2 0 Z (µm) Diameter (µm) Depth (µm) X (µm) Microstructuring of glass: fabrication and characterization CHAPTER 5 113 than 670ºC, the initial shape surface profile becomes flat, so microchannels obtained by laser direct-write tend to disappear. Figure 5.33 shows 3D confocal images of microchannels fabricated at different reflow temperatures. a) b) a) b) c) d) Figure 5.3 Confocal images of microchannels obtained after thermal treatment at different temperatures 620oC630oC 640oC650oC 660oC670oC Microstructuring of glass: fabrication and characterization CHAPTER 5 114 It can be appreciate in figure 5.33, that the diameter increase and the height decrease with the temperature. For a temperature of 630oC there is a smooth variation on shape, which is in concordance with the initial design, but the roughness is modified considerably, which is of great interest for microfluidics applications. The surface roughness is understood as an irregularity from a smooth surface caused during a fabrication process. These irregularities appear as microscopic holes and bulges from the expected (smooth) surface. The surface roughness can be defined as a vertical deviation between the highest bulge and deepest hole within a certain area (Fig. 5.34). Figure 5.34 Surface Roughness can be defined as an average vertical deviation between the highest bulges and lowest holes. An average surface roughness (for now on just surface roughness) can be specified with Eq. (5.3). (5.3) A smooth, optimal microfluidic channel has the specific resistance related to the geometry. The surface roughness on the wall of the channel increases by decreasing the flow rate, so the change of hydraulic resistance is proportional to the change of surface roughness. For determining the quality of the microchannels fabricated, we determine the roughness at the bottom of the channels. Table 5.4 shows the evolution of roughness with temperatures between 620ºC and 670ºC. Rz1 Rz2 Rz3 Rz4 Rz5 Microstructuring of glass: fabrication and characterization CHAPTER 5 115 Table 5.4 Comparative of roughness evolution with thermal reflow. Temperature Ra Ra Glass surfaces 3.68 nm Ra laser ablation 640 nm 620ºC 125.14 nm 630ºC 57.15 nm 640ºC 39.45 nm 650ºC 29.40 nm 660ºC 16.43 nm 670ºC 7.35 nm The initial roughness for unprocessed glass was 3.68 nm, after thermal process at 670ºC its decrease till 7.35nm, this value is near to optimal value for unprocessed glass although there is an important morphological change. The best result was obtained for 630ºC, it maintains the initial shape and a roughness of 57.15nm, which is a very good value for microfluidics applications on glass [Malalahalli2004]. Moreover the roughness average (Ra) at the bottom of the microchannels before thermal treatment was Ra = 640 nm and after thermal treatment at 670oC was Ra = 7.35 nm, this is in the order of the unprocessed sodalime glass surface, which means that the quality of surface is much better after thermal treatment. In order to fabricate different devices configurations, a study of the evolution of depth and diameter with the number of laser pass over the same place was done. To assess the evolution of the depth of the channels versus number of passes N, the edge of the trench is extracted using image processing after each pass, with successive passes superposed onto a composite image to study the influence of N while keeping the other laser parameters constant. Microstructuring of glass: fabrication and characterization CHAPTER 5 116 Figure 5.35 show the evolution of depth and diameter varying the number of laser passes. a) b) Figure 5.35 Evolution of a) depth and b) diameter varying the number of laser passes. It is evident from figures 5.35 that the channel aspect ratio α = h/d varies with the number of pass. The channel width d reaches its saturation value after 5 laser passes, increasing just around 200 nm per laser pass and varying only 1 micron after 5 passes, then is maintained almost constant. In contrast, the ablation height h changes dramatically with increasing number of pulses, reaching a saturation value after 6 pass, presenting an increase of 1 micron per laser pass, till it reach a height of 12 microns , after it maintains constant. This behavior can be related with the non-evacuation of debris generated during laser ablation from the bottom of the channel. For studying the microfludics capabilities of our technique, including the capability to form complex crosssectional shapes,we center our study on the fabrication of microchannels with different configurations. The main inconvenient of using a laser direct write technique for fabricating microchannels are in the junctions since the propelled material of the subsequent channels. is deposited on the existing microchannel. Figure 5.36 shows different examples of microstructures fabricated for microfluidic purposes. 0 1 2 3 4 5 6 7 8 9 10 11 12 13 0 1 2 3 4 5 6 7 8 9 10 Number of passes Depth( m) 13 0 1 2 3 4 5 6 7 8 9 10 Number of passes Diameter( m) 12 11 10 9 Microstructuring of glass: fabrication and characterization CHAPTER 5 117 Figure 5.36 3D Confocal images microchannels at different configurations. Microfluidics deals with the precise control and manipulation of fluids that are geometrically constrained to a small scale. These devices are used in microanalytical systems such as in capillary electrophoresis for DNA and protein separations [Khandurina2000, Liu2001, Bousse2003], as miniaturized channels for microfluidics[Stone2001, Whitesides2001], and as biochemical reactors and sensors [McCreedy2000]; future applications include cell biology and tissue engineering [Chiu2000].Typical methods for microstructuring glass includes electron beam lithography, photolithography, and wet and dry etching [Chen2001]. Although these techniques provides robust and high-quality microfluidic systems, the techniques used for fabrication are often time consuming and requires sophisticated equipment located in clean rooms. -7 60 60 120 180 240 180 120 0 0 Z (µm) Y (µm) X (µm) -1 -10 -4 -7 120 90 90 60 30 0 60 30 0 0.0 1.5 3.0 4.5 X (µm) Z (µm) Y (µm) 450 300 150 0 0 150 300 450 600 0.0 0.8 1.6 2.4 Z (µm) Y (µm) X (µm) a) b) c) d) Microstructuring of glass: fabrication and characterization CHAPTER 5 118 Fabrication of complex microfluidic devices on glass quickly, easily and inexpensively is a topic of great interest. In this context, the advantage of the process we propose comes from the simplicity and speed of the pattern transfer into the substrate, even in the production of more complex structures as demonstrated in figure 5.36. The ability to fabricate glass microfluidic systems using the technique presented in thesis makes glass attractive for many applications in which speed and economy of production are desired. CHAPTER 6 APPLICATIONS ON SURFACES MICROSTRUTURING In this chapter the structuring capabilities of microlens arrays for micro-patterning surfaces is presented. Consequences of the material processing over the microlens such as damage produced by the spelled material will be analyzed. Solutions to avoid the damage will be done, in particular the Talbot effect as a tool to overcome this problem will be presented. Applications on surface microstruturing. CHAPTER 6 120 6.1 Introduction and motivation. The modification of surfaces at micrometer and sub-micrometer scales is considered to be a key future technology. This has led in recent years to a huge interest in the generation of micro and nanometer structures on surfaces. Various forms and techniques of surface texturing were developed over the years [MingHsien2003, Sandip2009, Etsion2005, Duarte2008, Razzaque2008]. Among them, laser ablation has the advantage of great versatility, since it can be adapted to produce a wide range of structures. Lasers have also demonstrated utilization in rapid surface microstructuring and modification of materials to generate the textured surface features without significantly affecting or altering the properties of the bulk material. These lasers can therefore handle surface microfabrication on material substrates reliably, easily, and at high throughput. Most of the laser-assisted microstructuring studies deal with the selective ablation of material to produce micrometer and sub-micrometer features for diverse applications. Laser texturing of surfaces for enhancing tribological performance has demonstrated significant improvements in load capacity, wear resistance, friction coefficients, etc. In such tribological applications laser textured grooves or pockets act as lubricant reservoirs and increase the lubricant film thickness on the sliding surfaces [Dumitru2000, Eskin2005]. Recently, laser-assisted surface microstructuring of porous ceramic materials have attracted significant interests from the viewpoint of improving the surface properties by generating geometric surface features [Hao2003, Hao2004, Hao2004b, Hao2004c, Harimar2004]. To realize the full potential of the laserassisted surface microstructuring of ceramics for applications necessitating rapid and precision control of geometric features, the efforts have to be focused on the implementation of rapid manufacturing techniques Applications on surface microstruturing. CHAPTER 6 121 In this context, micro-optics plays an important role in various industrial production processes, like micro-electronics, laser micro-machining and materials processing. Microlens arrays offer a unique enabling technology in critical domains such as sensors, communications, metrology, and medical imaging, often providing solutions where other technologies prove unsuitable, unwieldy or cost prohibitive. The use of microlenses is particularly well suited for manufacturing, optimization of laser beam delivery and improvement of production efficiency with the intention of cost reduction. Is in this framework where we propose the use of microlens arrays for surfaces micro-patterning of materials. 6.2 Structuring capabilities of microlens arrays. In the work presented in this thesis it was demonstrated the influence of Sn impurities over the ablation process of soda-lime glass. These impurities led to a reduction of the ablation threshold of glass, which makes the soda-lime glass suitable for fabricating elements using relatively nanosecond low power lasers. On the other hand, when we use the microlens for microstructuring surfaces materials the main disadvantage arises also from the impurities since it is possible to damage the glass using not very high fluences. Another damage mechanism comes from any imperfection or debris generated during the surface texturization performed with the microlens array, which may impinge and be deposited at the microlens surface, and acts as absorbing center leading to the ablation of microlens array and its subsequent deterioration. In order to avoid the damage of the microlens the optimum laser parameters have been selected. Low energies per pulse (≈ 60 µJ) and high repetition rates (50 kHz) have been used, increasing the interaction time and the number of pulses needed to generate the hole. Although this can be improved by using low absorption and high quality optical materials, as the microlens work well at low Applications on surface microstruturing. CHAPTER 6 128 T z f n z being 2 2 T a z (6.1) where f is the focal length of the microlenses, n the order of Talbot plane, zTis the Talbot distance, a is the period of the microlenses, and λ the wavelength of the incident light. Figure 6.6 shows several the Talbot planes obtained at different distances from microlenses when illuminating the microlens array with a laser λ=630 nm). a) b) c) Figure 6.6 Image obtained at different distance from microlens arrays a) 510 µm, (focus of microlens) , b) at 15,14 mm (first talbot image, c) 31.65 mm (second talbot image).Laser wavelength 630 nm. By illuminating the microlens array with a laser beam, we can use the resulting Talbot images to structure surfaces with the advantage of creating a multistructuring with great versatility in design (due to the dependence of the pattern replicated with the Talbot plane being exploited) and increasing the working distance (distance between microlens array and substrate) in order to avoid deterioration of the microlenses caused by the particles spelled from the substrate, and therefore increasing its useful life. To this purpose we have developed a system for surface multi-structuring (presented in Fig.6.7). It consists of a Nd: YVO4 Q-switch that works with the Applications on surface microstruturing. CHAPTER 6 129 following parameters (λ = 1064 nm, pulse width 20 ns at 10 kHz, M2 <1.2) in combination with a galvanometer and a flat field lens of focal length 120 mm., The microlenses array is placed between the laser and the substrate. In this case we used microlens with diameter 60 µm to have a distance enough between the Talbot plane and the microlens array for avoiding array damage. The substrate, on which we perform the microstructure, is placed in the focal plane of the field lens, whereas the microlens array is positioned such that one of the Talbot planes matches on the substrate (see the setup in fig.6.7). Figure 6.7 Experimental setup for microstructuring a material using the Talbot effect. The laser parameters used in this case were: pulse repetition rate of 15 kHz and energy per pulse of 450 μJ (this energy level does not damage the microlens array). The Talbot plane selected for microstruturing (considering that in this case the laser wavelength is 1064 nm) corresponds to the second Talbot image, corresponding to a length of 14.70 mm from the microlenses. HAZ LASER ESPEJOS GALVANOMÉTRICOS LENTE DE CAMPO MATRIZ DE MICROLENTES SUBSTRATO Laser beam Flat-field lens Substrate Galvanometer system Microlens Talbot distance Applications on surface microstruturing. CHAPTER 6 130 Figure 6.8a and 6.8b shows one of the multistructuring results obtained on a Cr foil, placing the microlens array substrate at 14.70 mm of the sample, which corresponds to the second Talbot image. Figure 6.8c and 6.8d shows the microstrutring results by placing the Cr foil at the focal length of the microlens arrays (1.180 mm). The targets were irradiated with Nd:YVO4 lasers operating at 1064 nm. The laser parameters used were 30 kHz, 5 W and exposing time of 5 seconds. Figure 6.8 Cr foil microstrutured using the microlens array by a) Talbot effect and 1 pass of laser beam, b) Talbot effect and 2 passes of laser beam, c) placing the CR at focal plane of microlens and 1 pass and c) at focal plane and 2 passes. It can be appreciate in figure 6.8b and 6.8d that by applying two times the laser treatment, the structure obtained at surface of Cr foil makes more irregular due 160 200 240 280 50 80 110 140 X (µm) Y(µm) Z (µm) 40 100 160 220 160 200 240 280 50 80 110 140 X (µm) Y(µm) Z (µm) 40 100 160 220 a) b) c) d) Applications on surface microstruturing. CHAPTER 6 131 to the increase of holes diameter with the second pass. The Talbot effect allows us to increase the working distance of around 1 cm respect to the length used for microstructuring if we did not use a Talbot plane, thus avoiding any damage of the microlenses by material generated during the ablation process. Figure 6.9a shows a 3D confocal image of the microlens array after microstructuring materials using the Talbot effect and figure 6.7b a microlens array after microstruturing materials at the focal plane of microlens. a) b) Figure 6.9 a) 3D confocal image of the microlens array after microstructuring materials using the Talbot effect and b) a microlens array after microstruturing materials at the focal plane of microlens. It can be appreciate in figure 6.9a that using the Talbot effect the microlenses do not suffer damage at surface. On the other hand, working at focal plane (fig. 6.9b), the particles spelled from the target are deposited at surface of microlens, leading to the damage of microlens arrays. Applications on surface microstruturing. CHAPTER 6 132 CONCLUSIONS In this chapter the conclusion emerged from the works presented in this thesis are presented. Conclusions 134 Conclusions From the work realized in this thesis it has been emerged the next conclusions: A newmethod for fabricating microlens and microfluidic microchannels on soda-lime glass has been developed. It consists ofa combination of the laser direct write technique for fabricating the promoting glass structures, a wet chemical etching to remove the roughnessgenerated during laser ablation and a thermal treatment for reshaping and or improving of the optical and morphological qualities of the generated microlens and microfluidic microchannels. It has been showed the suitability of the developed method for glass microstructuring.. The use of pulsed nanoseconds lasers for the laser direct write, includes the benefits of using lasers commonly implemented for laser processing of materials applications, which makes the technique presented in this thesis highly competitive compared with other techniques commonly used on glass microstructuring. It has been demonstrated the use of IR nanosecond lasers as a tool for microstructuring glass materials, which due to the low absorption at IR, are typically machined with excimer laser or solid state lasers working in the UV wavelenght. Related with the chemical etching process, it has been revealed as a suitable process for removing the debris deposited over the glass after laser ablation, reducing the roughness andeliminating the secondary Conclusions 135 postunablated due to themicrolens packaging. Although the chemical etching has been analyzed for different microlens diameters, the best results were obtainedfor microlens of diameter 140 µm. In this case,the secondary posts generated during laser ablation, which are bigger in the case ofmicrolens of diameter 140 µm, interact with the material of the main cylindrical post during the material displacement produced during the thermal treatment, leading to an irregular distribution. It has been showed that without using the chemical etching it was not possible to obtaining microlens of diameters higher than 140 µm. For diameters below 140 µm does not present any significant result that justify the use of acid fluorhidric. The influence of the thermal treatment, which was applied at a range of temperatures between 620ºC and 670ºC, revealed two different ways to proceed, at high temperatureit reduces the roughness up tothe same level presented bythe glass before being processed, increasing the optical quality but with a significant change of shape (interesting for obtaining good microlens).At low temperatureit keeps the shape and decrease the roughness(appropiate for fabricating microchannels). It has been showed that when the thermal treatment was applied,the shape of the microlens evolves from a flat-top post to a quasi-parabolic profile as the temperature treatment increases. In the case of 620ºC some irregularities in the surface generated by the laser ablation step were observed. For temperatures of 630ºC and higher these irregularities disappear. The diameter of the microlenses is maintained almost constant over the range of temperatures tested, while the sag decreases due to the redistribution of the material from the top sides of the trench to the Conclusions 136 bottom of the crater. Temperatures out of this range were also analyzed; it was found that for values lower than 620ºC there were almost no significant changes in the surface shape. By contrast, for temperatures higher than 670ºC and 2 hours of thermal treatment, the initial array of cylindrical surface profile becomes flat, so the microlenses obtained by laser direct-write disappear. Using the laser-direct write technique, the chemical etching and the thermal treatment it was posible to fabricate microlens of diameters 40 µm, 60 µm and 140 µm with focal lengths of 510 m 5 m, 1.180 mm 0.050mm and 5.100 mm 0.050mm. We also fabricated microlens of 20 µm but in this case the focal length was not analized since the focus distribution were not uniform. It has been showed how the optical quality improves by increasing the reflowing temperature. This improvement is manifested in the reduction of the focal spot sizes, their homogeneity and the better irradiance distribution. The optical aberrations of the microlens arrays fabricated at 650ºC, 660ºC and 670ºC have been also determined. The topographic data obtained with confocal microscope have been used for analyzing the optical aberrations of the microlens array by calculating the modal coefficients of the expansion of the height data in terms of the orthogonal Zernike polynomial basis. The best result has been obtained for the microlens generated at 670ºC, which present a wavefront root mean square error of /28 and a variability of /77 Conclusions 137 Concerning microchannel fabrication the proposed techniques allowed us to obtain microchannels of a minimum diameter of 8µm and depth of 1.5 µm. A decrease of the roughness average generated after laser ablation,of two orders of magnitudereaching values of theorder of theunprocessed glass, has been obtained thanks to the thermal treatment. The laser-matter interaction process using different pulses width, covering the nanosecond, picosecond and femtosecond regimens has been investigated. The data obtained have provided useful information for a better understanding of the laser-matter interaction process, which was analyzed as a sequence of simpler interconnected problems: the absorption of laser light, the ionization and energy transfer from electrons to ions, the heat conduction and hydrodynamic expansion resulting in the ablation process. It has been showed that the ablative process in glasses can be promoted by: 1) impurities presented in soda-lime glass; 2) surface roughness; 3) pulse width; and 4) defects at surfaces (that acts like scattering center for the incoming light promoting its interaction with the surface of transparent materials). It has been identified, using EDX analysis, different percentage of Sn impurities on each side of the soda-lime glass. In order to evaluate the influence of these impurities as seed electrons for initiating the ablation process, the ablation threshold was identified independently over each side. The laser used was an IR nanosecond laser operating at 1064nm. The ablations thresholds identified correspond to 112 J/cm2 for the side A