This action might not be possible to undo. Are you sure you want to continue?
DISSERTATION Presented in Partial Fulfillment of the Requirements for The Degree Doctor of Philosophy in Graduate School of the Ohio State Univeristy
Hae Woon Choi, M.S. ***** The Ohio State University 2007
Dissertation Committee: Professor Dave F. Farson, Adviser Professor Charles E. Albright Professor L. James Lee ________________________________ Welding Engineering Graduate Program Approved by
Femtosecond laser ablation has interesting characteristics for micromachining, notably non-thermal interaction with materials, high peak intensity, precision and flexibility. In this dissertation, the potential of femtosecond laser ablation for fabrication of biomedical and electronic devices is studied. In a preliminary background discussion, some key literature regarding the basic physics and mechanisms that govern ultrafast laser pulse interaction with conductive materials and dielectric materials are summarized. In the dissertation work, results from systematic experiments were used characterize laser ablation of ITO (Indium Tin Oxide), stainless steel (hot embossing applications), polymers (PMMA, PDMS, PET, and PCL), glass, and fused quartz. Measured parameters and results include ablation threshold, damage threshold, surface roughness, single- and multiple-pulse ablation shapes and ablation efficiency. In addition to solid material, femtosecond laser light interaction with electrospun nano-fiber fiber mesh was analyzed and studied by optical property measurements. Ablation of channels in nano-fiber mesh was studied experimentally. Internal channel fabrication in glass and PMMA polymers was also demonstrated and studied experimentally. In summary, it is concluded that femtosecond laser ablation is a useful process for micromachining of materials to produce
microfluidic channels commonly needed in biomedical devices such as micro-molecular magnetic separators and DNA stretching micro arrays. The surface roughness of ablated materials was found to be the primary issue for femtosecond laser fabrication of microfluid channels. Improved surface quality of channels by surface coating with HEMA polymer was demonstrated.
Dedicated to God, my wife, my families and teachers
Dr. for all of her love. for providing me with the opportunities and support to achieve my goals. My mentor. help and encouragement. Chungnyun boo mokjang family were my greatest supporters in prayer. This dissertation is a result of much collaboration with other NSEC fellows. Dave Farson. Edison Welding Institute (EWI) and the Nano Science and Engineering Center (NSEC) program at OSU have been providing a comfortable research environment as well as their financial support in my life of graduate study. Dr. I would like to thank God for the successful completion of this dissertation. This dissertation is also dedicated to my parents. Yong Chae Lim. As committee members. I dedicate this work to my wife. I would like to thank to my advisor. Allen v . femtosecond laser for metallic materials. Dr. In chapter 3. for his invaluable advice. Charles Albright. Jin Sun. I had collaboration with Prof.ACKNOWLEDGMENTS There are many people whom I would like to thank for their support of my research at The Ohio State Univeristy. D study. Stanislav Rokhlin. Pastor Jay Chun encouraged me a lot during Ph. Dr. endless patience and support. I also thank to our lab colleagues. and Min Hyun Cho for their support and encouragement. Jian Chen. James Lee provided me a lot of references and research considerations. and my lovely brothers and sisters.
John Lannutti. Prof. I also thank to Nick Farrell for his support on characterization of channels and other valuable advice. We enjoyed the research together and shared much information. Susan Olesik supported me to complete the research. Chunhe Zhang and Chunmeng were dedicated to getting the results and I appreciate their time and effort. In chapter 4. Burr Zimmerman and Jeremy Steach are the men whom I should specially thank. vi . Lei Li. and Ruth Li supported my research by providing PCL fiber and the results of cell growth. femtosecond laser for dielectric materials. and Prof. James Lee. Prof. Sarah Drilling. Prof. Jed Johnson. Jeff Chalmers. James Lee.Yi and Prof.
16th. BSME. 1972 …………………………… Born – Taegu. D study at Ohio State University vii . Samsung Heavy Industries ……………………………. Engineer. Daehyun Tech 2000 – 2003 ……………………………… Eng. South Korea 2001. Engineer. Manager. South Korea 1992 – 1996 ……………………………. Keimyung University Taegu. Univ.2003 …………………………….VITA Oct. Laser Tech USA 2003 – present …………………………… Ph. Central Florida Orlando. MSME. South Korea 1989 – 1991 1995 – 2000 …………………………….
Applied Optics 46 (23). 2007 FIELD OF STUDY Major: Welding engineering Minor: Biomedical engineering viii . D F Farson. M Cho. D F Farson. 2006 2. 2006. H W Choi. D F Farson. D Ashkenasi. H W Choi. Journal Articles 1. 2006 4. 3. “A hybrid laser+GMAW process for control of the bead humping defect”. “Direct-write patterning of ITO film by high pulse repetition rate femtosecond laser ablation” . A Arai. S I Rokhlin. C Lu. D F Farson. J Bovatsek. Welding Journal 85 (8): 174S-179S Aug. “Electrical discharges between platinum nanoprobe tips and gold films at nanometer gap lengths” Nanotechnology. Aug. H W Choi. H W Choi. and L. 18 (3): 210-215 Aug.PUBLICATIONS A. 17(1):132-139 Jan. James Lee “Femtosecond bulk laser micromachining of microfluid channels in PMMA” Journal of Laser Applications.
.........................................TABLE OF CONTENTS ABSTRACT..........1 Femtosecond laser system............................................................... 2 Chapter 2 Femtosecond laser for micro/nano-scale machining.............................................................................................2 The mechanism of sub-picosecond ablation ............... v VITA ............................................................................................3................................................................................. 12 2.................................................................1 Micro-/nano-scale machining technologies ........... 8 2....................... ii ACKNOWLEDGEMENTS.................................................................. 16 2.... xvii Chapter 1..............4 References....................................................................................... 8 2.........................2 Organization of dissertation.............................................................................................................. 11 2.........................................................................................................2 Beam delivery system ..........................................................................................1 Introduction....................................................... 18 ix ....... 1 1............................. vii LIST OF FIGURES .......................................................... 12 2............................... 1 1.........................................................................................................................................................................................3 Femtosecond laser system for experiments ....3........ xiii LIST OF TABLES..........................................
Chpater 3 Femtosecond laser interaction with metallic materials .................................... 21 3.1 ITO ablation ................................................................................................................ 21 3.1.1 Introduction...................................................................................................... 21 3.1.2 Experiments ..................................................................................................... 27 3.1.3 Results and Discussion .................................................................................... 28 3.1.4 Conclusions and future works.......................................................................... 49 3.1.5 References........................................................................................................ 51
3.2 Hot embossing ............................................................................................................ 53 3.2.1 Introduction..................................................................................................... 53 3.2.2 Experiment procedure..................................................................................... 56 3.2.3 Laser ablation.................................................................................................. 61 3.2.4 Ablation and damage thresholds and pattern resolution ................................. 61 3.2.5 Mold micromachining..................................................................................... 74 3.2.6 Hot embossing ................................................................................................ 79 3.2.7 Conclusions and future works......................................................................... 85 3.2.8 References....................................................................................................... 87
Chapter 4 Femtosecond laser interaction with dielectric materials .................................. 90 4.1 Poly-caprolactone (PCL) machining........................................................................... 90
4.1.1 Introduction...................................................................................................... 90 4.1.2 Experimental Apparatus and Procedure........................................................... 94 4.1.3 Results.............................................................................................................. 98 4.1.4 Conclusions and future works........................................................................ 113 4.1.5 References...................................................................................................... 114
4.2 Laser machining of dielectric materials for biomedical applications ....................... 116 4.2.1 Introduction................................................................................................... 116 4.2.2 Analysis of Laser Ablation ........................................................................... 120 4.2.3 Experimental Apparatus and Materials......................................................... 123 4.2.4 Ablation Threshold Measurement................................................................. 128 4.2.5 Microchannel Ablation - Procedures and Results......................................... 130 4.2.6 Surface roughness and HEMA coating process............................................ 144 4.2.7 Surface roughness characterization............................................................... 147 4.2.8 Conclusions and future works....................................................................... 151 4.2.9 References..................................................................................................... 152
4.3 Femtosecond Laser Bulk Micromachining of Microfluid Channels in PMMA ....... 156 4.3.1 Introduction.................................................................................................... 156 4.3.2 Concept of gas assisted material removal:..................................................... 160 4.3.3 Experimental .................................................................................................. 162 xi
4.3.4 Results and discussion ................................................................................... 164 4.3.5 Conclusions and future work ......................................................................... 174 4.3.6 References...................................................................................................... 175
Chapter 5 Measurement of femtosecond laser scattering in ES-PCL nanofiber ............ 177 5.1 5.2 5.3 5.4 5.5 Introduction............................................................................................................ 177 Experiment setup ................................................................................................... 184 Results.................................................................................................................... 190 Conclusions and future works................................................................................ 196 References.............................................................................................................. 198
LIST OF FIGURES Figure 2.1 Schematic of seeding laser source .................................................................. 14 Figure 2.2 Regenerative amplifier .................................................................................... 15 Figure 2.3 External beam delivery system........................................................................ 17 Figure 3.1 Geometry of overlapping ablation spots ......................................................... 26 Figure 3.2 Single pulse ITO ablation energy-diameter curves. ........................................ 29 Figure 3.3 ITO squared ablation diameter vs. fluence...................................................... 30 Figure 3.4 Ablation depth of ITO using different focus optics ........................................ 31 Figure 3.5 Surface profile of laser ablation of ITO .......................................................... 32 Figure 3.6 Single-pulse ablation threshold calculation of glass substrate ........................ 34 Figure 3.7 2 kHz multipulse ablation................................................................................ 35 Figure 3.8 100 kHz laser ablation results ......................................................................... 36 Figure 3.9 Ablation line width vs. fluence........................................................................ 37 Figure 3.10 Variation of line ablation threshold with focus spot overlap ........................ 39 Figure 3.11 Geometry of overlapping ablation spots ....................................................... 43 Figure 3.12 Scan speed vs. pulse repetition frequency predicted from the results........... 45 Figure 3.13 Predicted and experimental ablation width versus scanning speed............... 47 Figure 3.14 Steps for fabrication of a polymer microfluid device.................................... 57 Figure 3.15 Surface profile of polished mold blank ......................................................... 58 Figure 3.16 Block diagram of beam delivery system setup.............................................. 60 xiii
.........4 Depth of grooves machined in ES nanofiber meshes......................... right) scans s............ fluence .7 (a) Focused and (b) defocused (multiple passes.................................22 Overlap of adjacent focus spot locations. 102 Figure 4...........5 Comparison of grooves ablated in (a) PCL and (b) PET ............. 73 Figure 3.............20 Ablation volume as a function of laser pulse energy ............................................. 82 Figure 3.24 (a) Hot embossing mold on stainless steel........................... (b) surface profile of mold ................6 Two grooves ablated in PCL nanofiber.............27 Theoretical diameter of ablation and damage regions.............. 68 Figure 3... 85 Figure 4. 99 Figure 4......... 83 Figure 3................... Right E=65nJ) ...............................21 Ablation efficiency ................19 3D profile of laser ablation (Left E=500nJ....... 106 Figure 4...................................28 Ablation profiles for pulses with various peak fluence.........23 Multipulse overlapped scanning (E=165nJ)....2 SEM images of (a) Q-switched laser and (b) femtosecond laser grooves. 97 Figure 4. 71 Figure 3... 77 Figure 3........ 103 Figure 4..............17 Single pulse ablation patterns for various pulse energies... 84 Figure 3............... 105 Figure 4.............................................1 Fiber diameter size distribution of ES-PCL used in the experiments ..26 SEM images of an embossing tool machined ..............................................................8 Groove width squared of nanospun polymer vs...............................3 Width of grooves machined in ES nanofiber meshes...................... 108 xiv ...18 Ablation diameter results with regression-fit lines....Figure 3............................. 65 Figure 3....25 Image of hot embossed molecular magnetic separator pattern in PMMA ... 80 Figure 3............. 70 Figure 3.......................... 104 Figure 4.................................. 66 Figure 3...............
..........22 Optical images of the channel quality and shape .24 Optical images of HEMA coated channels (before and after) .. (c) 50%.... 146 Figure 4........................................... 124 Figure 4........ 142 Figure 4......................20 Accumulated fluence as a fraction of single pulse ............ (b) fluence............................. (d) 75%.........................................Figure 4.......................... 122 Figure 4.14 Single pulse ablation diameter (a) pulse energy.......................... 146 Figure 4.................................................... overlap............................... 129 Figure 4.... 110 Figure 4.............13 Block diagram of beam delivery system setup..............................................................16 Ablation depth as function of pulse energy...12 Definition of focus spot overlap LO.... 140 Figure 4....... 118 Figure 4..........19 Accumulated laser fluence (a) 0%.... 148 Figure 4........... 143 Figure 4................ 136 Figure 4...........17 Profilometer scan across channels ablated at pulse energy and pitch .....5 µm..........10 Matrix of microscale structures ablated in PCL nanofiber mesh .... 138 Figure 4.....18 Peak-peak surface roughness versus pulse energy at a pitch of 2................. 149 Figure 4............27 Surface roughness data for HEMA coating...11 Keldysh parameters for soda lime glass and fused quartz.....25 Surface roughness data for soda-lime glass .........................................9 (a) transmittance of ES PCL (b) reflectance of nanofiber mesh .....28 Surface roughness data of soda-lime glass and fused quartz ............15 Ablation pattern with respect to laser fluence and scanning speed .............. (b) 25%...................... 112 Figure 4. 132 Figure 4..21 Fluence variation vs. 133 Figure 4...23 HEMA coating process .. ............ 148 Figure 4.............. 150 xv .............. 134 Figure 4.........................26 Surface roughness data for fused quartz.................................................
....... 160 Figure 4........................ 169 Figure 4...2 Definition of TE and TM modes ......7 Transmittance measurement for fiber PCL and solid PCL ...... 192 Figure 5............4 System setup for transmittance and reflectance measurements ..29 Concept of gas assisted material removal .... 170 Figure 4................. 185 Figure 5................................ 196 xvi ...............6 ml/s .33 Plan view of channels created with multiple passes.................................................. 180 Figure 5............................................Figure 4......................................65 ml/s...............3 Electro spun fibers (left) and fiber diameter distribution (right).......... 181 Figure5......................................................................... 189 Figure 5............................30 Experimental setup ....... 188 Figure 5.................. 194 Figure 5............5 System setup for transmittance and reflectance measurements .............10 Increased ablation diameter by multiple scattering . 1.................................................................... 165 Figure 4......................................36 Micrograph of internal channels before being filled with a liquid.......11 Ablated fiber with femtosecond laser (E=500nJ).. 167 Figure 4........ 195 Figure 5.. 163 Figure 4.................35 Reynolds numbers in the flow channel for 0..................... 2..........9 Ablation diameters vs......................................................34 Cross sectional view of channels.....8 Reflectance measurement....31 Channel fabrication without gas assistance................. 172 Figure 4............32 Channel fabrication with gas assistance at different flow rates ..6 Output power vs... 191 Figure 5.............37 Flourescense spectra confirming presence of yellow dye in a channel........ 171 Figure 4....1 Light propagation inside and outside of a fiber mesh ........ Fluence ............................ 190 Figure 5................. 173 Figure 5.3 ml/s............... input power of samples......................
................................................ 91 Table 4.....................................1 Single pulse ablation thresholds and incubation factors.................1 Physical properties of PCL and PET ..................LIST OF TABLES Table 2..............1 Material properties of PCL ........................... 186 xvii ........... 13 Table 3......................1 Laser systems in Femtosecond laser..........................................................................................................3 Material properties of fused-quartz............ 128 Table 4. 41 Table 4...........4 Experimental parameter settings..............................................................................2 Material properties of soda-lime glass....... 131 Table 5............ 127 Table 4......................
and non-clean room operation provide many advantages in micromachining for biomedical applications. 4] and a wide range of materials can be machined. non-thermal interaction. At this high intensity.CHAPTER 1 INTRODUCTION 1.1 Introduction The femtosecond laser has opened a new era of micromachining. The main characteristic of femtosecond (or ultrashort) laser pulses is very high peak intensity (> 1016 W/cm2) and rapid deposition of energy into the material. Femtosecond lasers with pulse durations of ~100fs have been widely used for micromachining because of collateral damage is mostly eliminated in dielectric materials  and heat affected zones in metal can be smaller . high peak power. Benefits of ultrashort laser pulses arise because the pulse duration τ ~ 150fs is less than the time τei 1–10 ps equilibration between electron and lattice-ions subsystems. Its precision. the onset of optical breakdown that initiates ablation can be more deterministic than stochastic [3. 1 . flexibility.
degrading the quality of machined surfaces and shortening the life of machine tools. The short duration can eliminate the affects of hydrodynamic motion of the matter during laser irradiation . Ti-Ni alloys undergo severe strain hardening during mechanical machining. In addition. Due to these advantages. Also. The applications of femtosecond laser ablation that we developed during this research are discussed in following chapters. the time τh needed for the electron thermal energy diffusion to reach the optical penetration depth is several orders of magnitude longer than τ . 2 . stress-induced martensitic transformation at the cutting front leads to an undesirable contact interface between the tool and the workpiece . However. femtosecond laser provides a very flexible way to fabricate devices of nearly all types of materials. femtosecond laser was proven to be a very effective processing tool for such hard materials and a wide variety of others as well. For example. Many hard metals or carbides which would make very rugged and durable embossing molds are not compatible with micromachining by etching processes. The short pulse also can prevent the shielding of incoming laser beam by plasma from the ablated surface. the same superior mechanical properties also create difficulties for mechanical machining. which can maximize absorption efficiency .In addition.
The single-pulse ablation threshold of ITO was smaller than that of the glass substrate so the entire thickness of ITO could be removed in a single pulse or with multiple intersecting pulses without the possibility of substrate ablation.2 Organization of dissertation This dissertation is organized into 6 chapters including this chapter.1. A broad range of laser energy. A systematic study of ablation threshold. contributing to large feasible scanning speeds for high pulse frequency lasers. stainless steel grade 304L. Chapter 2 is prepared to provide an introduction to the fundamentals of femtosecond lasers and the processing systems that were setup for the experiments. Laser pulse energy and focus spot size were varied in single-spot ablation tests and for ablation of linear features with scanned multiple pulses. ablation efficiency and affect of machining parameters on mold accuracy was performed. An analysis showed that incubation affects lowered ITO ablation thresholds when pulse frequency was high relative to scanning speed. 13nJ to 3 . Chapter 3 provide experimental results for the laser micromachining of conducting materials and applications such as indium doped with tin oxide (ITO) film for digital displays and stainless steel for hot embossing of polymer. The first part of the chapter describes ITO ablation. Femtosecond laser micromachining was used to fabricate hot embossing molds of a durable material. The second part of Chapter 3 describes a hot embossing mold fabrication with femtosecond laser on stainless steel. Linear features could be created at much higher scanning speeds using a high repetition frequency (100 kHz) Yb fiber amplified laser as compared to a lower repetition frequency (2 kHz) laser.
In addition.2%. 75% pulse overlap and 135nJ laser power was found to be the optimum process with minimum recasting. some melting of fibers was observed at the edges of grooves. Also. pulse energy and scanning speed were varied to determine their affects on groove size and the characteristics of the electrospun fiber at the edges of these grooves. Chapter 4 describes the laser beam interaction with dielectric materials such as glass. A finished mold was used was then used to fabricate a microfluidic micromolecular magnetic separator in poly-methyl macryolate (PMMA) polymer. For scanned ablation. a novel method to fabricate internal microfluidic channels by femtosecond laser is presented. Also. Nano fiber meshes of PCL and PET and PET were structured by ablation of linear grooves with a scanned femtosecond laser. and electrospun poly-caprolactone (ES PCL) and poly-ethylene terephthalate (ES PET) fiber mesh. and the most effective region of ablation was found to be between 65nJ and 135nJ. 4 . The femtosecond laser was seen to be an effective means for flexibly structuring the surface of ES PCL scaffolds. was applied. the width of the ablated grooves was well-controlled by laser energy and focus spot size although the grooves were significantly larger than the spot size. fused quartz.500nJ. Femtosecond ablation resulted in much more uniformly ablated patterns compared to Q-switched nanosecond pulse laser ablation.189J/cm2 and ablation efficiency was estimated to be around 1. The first section describes PCL and PET ablation. These affects were attributed optical radiation and multiple scattering from laser-induced plasma at higher pulse energies and melting of fibers at laser fluences lower than the ablation threshold. The ablation threshold for stainless steel was found to be 0. Focus spot size.
Internal channels in polymer are widely used in biotechnology applications such as DNA stretching and in devices such as Micro Total Analysis Systems (μ-TAS) and Lab on Chip (LOC) systems. resulting in the ablation of an internal channel. soda-lime glass surfaces were smoother than fused quartz. we describes techniques for microfluidic channel fabrication in soda-lime glass and fused quartz using femtosecond laser ablation and ablation in conjunction with polymer coating for surface roughness improvement. For some applications. At high fluence.In the second part. smoother channels are desired. femtosecond pulsed laser energy has been used to implement a convenient direct write bulk-machining process in glass. Laser fluence and focus spot overlap showed the strongest influence on channel depth and roughness. and material properties) affect machining quality for dielectric materials. In this technique. scanning speed. For the same laser ablation parameters. roughness decreased to 100nm-350nm RMS and showed a greater dependency on overlap. At low fluence. the laser beam is focused inside of a transparent material. Systematic experiments were done to characterize how process variables (laser fluence. and focus spot overlap. For manufacturing prototype devices. Initial experiments for internal channel fabrication in poly-methyl methacrylate (PMMA) 5 . The surface roughness of laser ablation was also dependent on the material properties. The third part of this chapter describes a novel method to fabricate a microfluidic channel inside the transparent material. especially those using quartz. the surface roughness was measured to be between 395nm-731nm RMS. A hydroxyethyl methacrylate (HEMA) polymer coating was applied and roughness of coated channels was improved to 10-50nm RMS.
Relatively smooth channels with a minimum diameter of 2 μm. Soc. Ablation of metals by ultrashort laser pulses. Tunnermann. C.polymer revealed a significant problem with clogging of channels by debris and rough. we describe a new method to fabricate internal channel in PMMA using femtosecond pulsed laser energy and a gas-assisted material removal concept. Chichkov. B. 2716-2722. 1997. B. Physical Review Letters. Laser-induced damage in dielectrics with nanosecond to subpicosecond pulses. 14(10): p. and M. B. Nolte.3 References 1. J. Wellegehausen. 6 . M. a maximum diameter of 20 μm and a maximum length of 10 mm were achieved with this technique. In this paper. Opt. Multiscattering theory. Stuart.094 J/cm2 and it is observed that 98% of laser beam which is absorbed inside the mesh is scattered. B. A. Perry. Rubenchik. The ablation threshold is measured to be 0. 1995. H. S. 2. The light absorption in fiber mesh is different from that in solid material and scattering plays a significant role. Integrating sphere was used to measure the transmittance and reflectance of light in ES fibers. Welling. modified from Kubelka-Munk theory. B. 74(12): p. Jacobs. Momma. Chapter 5 is describing the measurement of optical properties for electrospun PCL nano fiber mesh. fractured channel walls. is used to quantify the scattering and absorption coefficients. Shore. Am. Feit. A. and H. 2248-2252. 1.
and M. J. S297-S301. 7 . X. Ph. 6. Proceedings of symposium H on photonic processing of surfaces . Sentis. T. 2004. D. Hermann. Lett. G. Smart Materials and Structurs. Squier. Laser induced breakdown by impact ionization in SiO2 with pulse widths from 7ns to 150fs. 1997. 1994. Itina. 4. 64(23): p. Mourou.Thin films and devices. H.Z. Mourou. Experimental investigations machiniability of Ni50. X. Du.3. Korn. 1706-1716.454. 2005. J. 3071-3073. Delaporte. 453 . Modeling of metal ablation induced by ultrashort laser pulses. and G. 14: p. D.. 5. Laser ablation and micromachining with ultrashort laser pulses. Du. and G.4.. and Y.6Ti49. H. Huang. Liu. Liu. Appl. E-MRS: p. Phys. IEEE journal of quantum electronics. 33(10): p. Liu.
One of the most popular fabrication technologies is a process technique typically used in semiconductor fabrication: photolithography. The basic configuration of femtosecond laser system and experiment setups used for this dissertation and are presented. Photolithography transfers a pattern from a mask to a photosensitive material using a radiation source such as mercury lamp or ultra violet (UV) laser. 2. a wide range of micro-/nano-scale fabrication techniques that are available both in clean room and non-clean room environments are overviewed.and nano-scale machining is becoming increasingly important for biomedical and semiconductor applications.1 Micro-/nano-scale machining technologies Micro. The femtosecond laser process is compared to other processes and its advantages are discussed based on the review of previous publications.CHAPTER 2 FEMTOSECOND LASER FOR MICRO/NANO-SCALE MACHINING In this chapter. By 8 .
The integration of conventional and AFM-based lithographic techniques brings the resolution of fabrication technology to a nanometer scale [10. Other techniques used include micromachining of silicon for microelectricalmechanical systems (MEMS) and embossing of thermoplastic with patterned quartz  Instead of a lithographic approach. Microtransfer Molding. many advanced micromachining techniques have been developed and used for a variety of applications [1-3]. and Microcontact Printing. Several different techniques are collectively described as soft lithography: Near-Field Phase Shift Lithography. The AFM with silicon or diamond tip etching 9 . Soft lithography is an alternative method to silicon-based micromachining and a nonphotolithographic strategy based on self assembly that uses replica molding of nontraditional elastomeric materials and bimolecules to fabricate stamps. 11]. Micromolding in Capillaries. Solvent-assisted Microcontact Molding (SAMIM). effective. The reported advantage for AFM is that the machining scale can be as small as the size of the AFM probe tip.and nano. Replica Molding.scale polymeric devices from a master mold by photolithography is called soft lithography. However. it still requires multiple steps from the development of appropriate masters to a working device and residual chemicals remaining after the process that can be toxic if implanted in a biomedical device . It provides a convenient. and low-cost method for patterning [4-8]. the atomic force microscope (AFM) has also been used for nano-scale machining. A different type of lithography used to fabricate micro.using a combination of both photolithography and polymer processing.
such as polymers to relatively hard layers such as oxide thin films [1. UV (Ultra Violet) laser such as Excimer lasers have been used as a micromachining tool on wide bandgap dielectric materials. This approach provides a flexible way to fabricate complicated geometry in non-clean room environment. and copper vapor lasers have been used [17-23]. BaF2. However. 16] As a non-contact and non-clean method. corrosive gas handling and UV radiation damages the optics made for the UV pulsed laser is still an issue and less desirable tool for industrial applications. 10 . especially in dielectric materials. a variety of laser systems such as Nd:YAG. With the photochemical interaction of the laser with target material. Excimer. 10-14]. several drawback to AFM etching is that it is a slow process to create sub-micron features by scratching the surface of the material with a nano-sized pyramidal tip. 24]. 20. the AFM tip wearing. 22. CO2. and quartz for optical components such as gratings and lens. However. it can eventually achieve very fine feature sizes[19.or micromilling machine can provide a way to fabricate sub-micron features by scratching the surface of a soft material with its sharp tip. Ion milling and reactive ion etching are another choice of micro-/nano-scale fabrication [15. In addition to the above process. and the problem of fabrication area needs to be solved . Due to the high absorption properties of the materials. such as CaF2. sapphire. The range of materials that have been studied for soft materials.
2 The mechanism of sub-picosecond ablation Femtosecond lasers have been used for material processing for less collateral damage in dielectric materials  and reduced heat affected zones in metal .2. the femtosecond laser can also prevent the plasma shielding of incoming laser beam by ablated material which can lead to maximum absorption efficiency . It is well known fact that the principle characteristic of femtosecond pulses. 11 . Moreover. The major benefits of an ultrashort laser pulse include its ability to produce very high peak intensity and rapid deposition of energy into the material. In addition. this can eliminate the hydrodynamic motion of the matter during laser irradiation and no fluid dynamics is involved during the laser and matter interaction . is that the duration of pulse is shorter than the time τei 1–10 ps for equilibration between electron and lattice-ions subsystems. The theoretical study of the transient evolution of the distribution function of the electron gas in a metal during and after irradiation with a subpicosecond laser pulse of moderate intensity has been studied by many prior researchers [29-32]. which makes process advantageous for metal treatment. By having a short duration. the duration of femtosecond pulse . the time τh needed for the electron heat diffusion to reach the optical penetration depth is several orders of magnitude longer than τ.
pulse duration less than 500fs and a higher repetition frequency of 100kHz and average power of 0.1 Femtosecond laser system In our experiments. The frequency and power are listed in Table 2.6 W. and Ti:Al2O3 laser for regenerative amplifier. CPA 2161 (Clark-MXR). This yields output at a wavelength of 1045 nm.2W The system setups are slightly different depending on experiments. The third system is based on an Er fiber oscillator with pulses amplified in a ytterbium (Yb) fiber-based amplifier followed by chirped pulse amplification in a Yb fiber-based amplifier. and FCPA μJewel D-400 (IMRA America) were used as femtosecond laser energy sources. The system has a single-mode erbium (Er) fiber oscillator to provide a seed pulses for chirped pulse amplification in a Ti:Al2O3 –based regenerative amplifier. and details will be explained at each section. adjustable from 3-6kHz. The CPA system consists of four lasers including a diode laser for Er fiber laser pumping source. pulse duration of 150 fs. CPA 2110 (ClarkMXR). commercial femtosecond laser systems. The first one has 775nm central wavelength.2.3. ORC1000 laser for Ti:Al2O3 laser pumping. SErF fiber laser which is a frequency doubled fiber seeding laser. pulse repetition frequency of 2kHz and an average power of 1.3 Femtosecond laser system for experiments 2.1 12 . The second system is similar to the first but has higher pulse repetition rate.
output coupling and wavelength control. 13 . some bulk optics used for polarization control. an active fiber ring laser. It is a unidirectional. and a temperature-stabilized periodically poled lithium niobate (PPLN) frequency doubler. The SErF contains a laser diode and associated control electronics. SErF is depicted in Figure 2. The pump source for the laser is a solid-state fibercoupled laser diode operating at approximately 980 nm. polarization rotation additively pulse mode-locked (APM) fiber laser which uses Erbium doped fiber as the gain medium. a compressor to eliminate residual dispersion of the output pulses. The schematic of the seed pulse laser.1 Laser systems in Femtosecond laser  SErF is based on the fiber ring “Stretched Pulse” laser.Laser Nd:YAG Ti:Al2O3 ErF Fiber SErF Fiber Wavelength (nm) 532 775 1550 775 Power < 65W < 5W 50mW < 20mW Pulse Energy < 25mJ < 5mJ < 10nJ < 1nJ Table 2.1.
The optimum injection or cavity dumping time is controlled by electronic devices (DT-505.Output Collimator Collimator WP BRF Isolator WP Erbium doped fiber Diode Laser WDM Figure 2.1 Schematic of seeding laser source  (WP: λ/4 waveplates.2. BRF: birefringent filter. The stretched seed pulse is then transferred to regenerative amplifier to amplify the laser intensity. The injected seed laser is amplified through Ti:Al2O3 cavity which is pumped by a frequency doubled Nd:YAG laser (λ=532nm) through dichroic through mirror shown in Figure 2. the polarization of a Pockels cell is changed to eject the amplified beam to the compressor. 14 . After multiple passes of amplification. Clark-MXR) and is very critical for the power and pulse duration. WDM: wavelength-division multiplexing) The seed pulse from the frequency doubled SErF laser (λ=775nm) is stretched to protect optics during pulse amplification. typically 5 times.
Figure 2. PC: Pockels cell. which is comparable 15 . RF: Radiofrequency)  The amplified laser beam then travels through compression optics where the width of pulse is compressed to back to the sub-picosecond region. This has been the main drawback of the process. Due to limitations of components used for optical pulse amplification. However. recent development of technologies enabled design of systems with pulse repetition rates in the MHz regime.2 Regenerative amplifier (HR: high reflective mirror. which in turn limited the maximum micromachining speed to inches/sec. CPA femtosecond laser pulse repetition rates remained around several kHz for some time. DM: Dichroic mirror. typically less than 150fs. FR: Faraday rotator.
the horizontally polarized laser beam is attenuated thorough the combination of thin-film polarizing beam splitters (PBSs) and a λ/2 wave plates. a series of polarizing beam splitters (PBSs) were installed to attenuate the power for micromachining. As shown in Fig. The direction of polarization is horizontal which is parallel to the optical table and set by the manufacturer.6W at 3kHz / 6kHz setting and 1. During the process.2Beam delivery system The gated pulse train produced by the laser systems has constant average power of 2.3.acoustooptically Q-switched pulsed Nd:YAG laser systems. Since the typical range of power for micromachining is less than 5mW.3. so laser maintenance is minimized. 2. a great amount of attenuation is required.6W at 1kHz/2kHz setting. The advanced technology decreased optical alignment requirements. 16 . 2.
3 where a perfect Gaussian beam is M2=1.5μm resolution and 25mm travel distance.Figure 2. The diameter of raw beam was measured to be 5mm and the beam quality.5μm resolution and 50mm travel distance. (LS055. A 50x infinity corrected microscope objective lens with 17 . The focusing optics is mounted on Z axis with 0. NMlaser Inc) and is finally delivered to the target material on a high precision XY stage (Parker-Hannifin) with 0. The laser beam can be turned on or off using an external high speed mechanical shutter with 4ms shutter control time. was measured by using autocorrelator to be 1.3 External beam delivery system The PBS regulates the transmission and reflection rate based on the polarization direction which can be adjusted by a λ/2 wave plate and incidence angle of laser beam. M2.
numerical aperture of NA=0.42 (50x M Plan Apo NIR, Mitutotyo) was used for fine focusing. For the consistency and easy focusing purpose, a coaxial vision system was installed and one can visually locate the material at the focus within +/- 1μm range where focal depth is 1.6μm for this selected optics.
1. C. Aguilar, Y. Lu, S. Mao, and S. Chen, Direct micro-patterning of biodegradable polymers using ultraviolet and femtosecond lasers. Biomaterials, 2005. 26: p. 7642-7649. J. Narashimhan, a.I.P., Polymer embossing tools for rapid prototyping of plastic microfluidic devices. J. Micromech. Microeng, 2004. 14: p. 96-103. F. Tay, J.K., F. Watt, and W. Choong, A novel micro-machining method for the fabrication of thick-film SU-8 embedded micro-channels. J. Micromech. Microeng, 2001. 11: p. 27-32. M. Unger, H.C., T. Thorsen, A. Scherer, and S. Quake, Monolithic microfabricated valves and pumps by multilayer soft lithography. Science, 2000. 288: p. 113-116. Y. Xia, and G. Whitesides, Soft lithography. Annu. Rev. Mater. Sci., 1998. 28: p. 153-184. J. McDonald, D.D., J. Anderson, D. Chiu, H. Wu, O. Schueller, and G. Whitesides, Fabrication of microfluidic systems in PDMS. Electrophoresis, 2000. 21: p. 27-40. R. Kane, S. Takayama, E. Ostuni, D. Ingber, and G. Whitesides, Patterning proteins and cells using soft lithography. Biomaterials, 1999. 20: p. 2363-2376. G. Whitesides, E.O., S. Takayama, X. Jiang, and D. Ingber, Soft lithography in biology and biochemistry. Annu. Rev. of Biomedical Engineering, 2001. 3: p. 335-373.
9. 10. 11. 12. 13. 14. 15.
G. Whitesides, Soft Lithography. 1998, WTEC hyper-librarian. T. Fang, C. Weng, and J. Chang, Machining characterization of the nanolithography process using AFM. Nanotechnology, 2000. 11: p. 181-187. Y. Kim, and C. Lieber, Machining Oxide Thine Films with an AFM: Pattern and Object Formation on the Nanometer Scale. Science, 1992. 257(5068): p. 375-377. J. Park, D.L., N. Takano, and N. Morita, Diamond tip cantilever for micr/nano machining based on AFM. Materials Science Forum, 2006. 505: p. 79-84. B. Bhushan, and V. Koinkar, Nanoindentation hardness measurements using AFM. Appl. Phys. Lett., 1994. 64(13): p. 1653-1655. Z. Hu, S. Zhang, and X. Zheng, Three dimensional micromachining based on AFM. Key engineering materials, 2006. 315: p. 800-804. E. Chanson, S.P., J. Poate, J. Borland, M. Current, T. Rubia, D. Eaglesham, O. Holland, M. Law, C. Magee, J. Melngailis, and A. Tasch, Ion beams in silicon processing and characterization. J. Appl. Phys., 1997. 81(10): p. 6513-6561. I. Adesida, A.M., E. Andideh, M. Khan, D. Oslen, and J. Kuznia, Reactive ion etching of GaN in silicon tetrachloride plasmas. Appl. Phys. Lett., 1993. 63(20): p. 2777-2779. H. Klank, J.K., and O. Geschke, CO2 laser micomachining and back end processing for rapid production of PMMA based microfluidic systems. Lab on a chip, 2002. 2(4): p. 242-246. B. Seddon, Y. Shao, J. Fost, and H. Girault, The application of excimer-laser micromachining for the fabrication of disc microelectrodes. Electrochimica Acta, 1994. 39(6): p. 783-791. J. Ihlemann, a.B.W.-R., Excimer laser micro machining of inorganic dielectrics. Applied surface science, 1996. 106: p. 282-286. G. Kopitkovas, T.L., C. David, A. Wokaun, and J. Gobrecht, Surface micromachining of UV transparent materials. Thin solid films, 2004. 453: p. 3135. A. Glover, E. Illy, and J. Piper, High speed UV micro machining of polymers with frequency doubled copper vaport lasers. IEEE journal of quantum electronic, 1995. 1(3): p. 830-836.
K. Naessens, H.O., P. Daele, and R. Baets, Flexible fabrication of microlenses in polymer layers with excimer laser ablation. Applied surface science, 2003. 208209: p. 159-164. E. Illy, D.B., M. Withford, and J. Piper, Enhanced polymer ablation rates using high-repetition-rate ultraviolet lasers. IEEE journal of quantum electronic, 1999. 5(6): p. 1543-1548. H. Endert, R.P., and D. Basting, Excimer laser: A new tool for precision micromachining. Optical and Quantum electronics, 1995. 27(12): p. 1319-1335. X. Liu, D. Du, and G. Mourou, Laser ablation and micromachining with ultrashort laser pulses. IEEE journal of quantum electronics, 1997. 33(10): p. 1706-1716. B. Stuart, M. Feit, A. Rubenchik, B. Shore, and M. Perry, Laser-induced damage in dielectrics with nanosecond to subpicosecond pulses. Physical Review Letters, 1995. 74(12): p. 2248-2252. S. Nolte, C. Momma, H. Jacobs, A. Tunnermann, B. Chichkov, B. Wellegehausen, and H. Welling, Ablation of metals by ultrashort laser pulses. J. Opt. Soc. Am. B, 1997. 14(10): p. 2716-2722. T. Itina, J. Hermann, Ph. Delaporte, and M. Sentis, Modeling of metal ablation induced by ultrashort laser pulses. Proceedings of symposium H on photonic processing of surfaces - Thin films and devices, 2004. E-MRS: p. 453 - 454. B. Rethfeld, A. Kaiser, M. Vicanek, and G. Simon, Ultrafast dynamics of nonequilibrium electrons in metals under femtosecond laser irradiation. Physical Review B, 2002. 65: p. 214303-1 - 214303-11. F. Vidal, T.J., S. Laville, O. Barthelemy, M. Chaker, B. LeDegoff, J. Margot, and M. Sabsabi, Critical-point phase separation in laser ablation of conductors. Physical Review Letters, 2001. 86(12): p. 2573 -2576. N. Bulgakova, Possibility of rarefaction shock wave under short pulse laser ablation of solids. Physical Review E, 1999. 60(4): p. R3498 - R3500. E. Gamaly, A.R., B. Luther-Davies, and V. Tikhonchuk, Ablation of solids by femtosecond lasers: Ablation mechnasim and ablation thresholds for metals and dielectrics. Physics of plasma, 2002. 9(3): p. 949-957. Clark-MXR, CPA 2110 User manual. 2 ed. 2004: Clark-MXR.
FEMTOSECOND LASER INTERACTION WITH METALLIC MATERIALS
In this chapter, femtosecond laser micromachining of conductive materials is presented. The first part describes thin film ablation with femtosecond laser. Indium oxide dope with tin oxide (ITO) is conductive transparent metallic layer which is used in display applications such as liquid crystal display (LCD), organic light-emitting diode (OLED), and other flat panel display application. The second part of this chapter describes a fabrication of a hot embossing mold for microfluidic channel applications.
3.1 ITO ablation 3.1.1 Introduction
Indium oxide doped with tin oxide (ITO) provides high electrical conductivity and transparency in the visible and near IR (infrared) wavelengths. Thus, it is widely used as a transparent electrode for the fabrication of liquid crystal displays (LCD’s) and organic light emitting diode displays (OLED’s), photovoltaic devices and other optical
for device manufacturing and is also needed for repair applications. Unique and generally desirable ablation characteristics are observed when irradiance is large and pulse length is shorter than the picosecond-scale times required for transfer of absorbed energy from electrons (either initially free or released from bound states by multiphoton absorption or tunneling) to the material lattice . Results pertinent to this investigation are summarized here.7. ITO is applied as a continuous film and is then locally removed to form conductors. thermally-induced defects. determined by composition. deposition process and parameters. In general. Its optical properties. can be minimized in the remaining material. Nd:YAG and Nd:YVO4 (fundamental and frequency-converted) [3.applications . many materials are ablated only when fluence exceeds a distinct threshold level. short pulse laser ablation offers intriguing possibilities for many microelectronics fabrication tasks [6.and nanosecond-pulsed laser ablation of ITO and multilayer transparent films have been reported . Consequently.4] and excimer  lasers have been used for ablation. one may consider single pulse removal of a material with ablation threshold fluence Fth by a beam with Gaussian radial fluence profile having a maximum greater 22 . in lieu of conventional photolithography. When illuminated with laser pulses of varying fluence. are important for applications and also affect laser processing . During device fabrication. device interconnections and other elements. Direct-write laser ablation removal of ITO from glass or other underlying materials is sometimes used. The femtosecond laser ablation has been analyzed  and results for femtosecond. which are often difficult to avoid with longer laser pulses. To analyze the size of ablation spots.8].
the radius at which the laser beam fluence equals the ablation threshold fluence can be found by solving the equation 2 Fth = F0 exp − 2r 2 w0 ( ) (3. (3. related to pulse energy Ep by 2 F0 = 2 E p / πw0 .1) to solve for the diameter D of the area near the center of the beam from which material is removed results in the relation 2 2 D 2 = 2w0 ln(F0 / Fth ) = 2w0 [ln (F0 ) − ln(Fth )] (3.2) Similarly 2 Fth = 2 Eth / πw0 . (3.1. single pulse ablation spots with measured diameter D are produced by varying pulse energy over a range of values while focus spot radius is 23 .1. Then.1.1.than the ablation threshold fluence.4) Experimentally. (3.1) where w0 is Gaussian beam radius and F0 is maximum (or peak) fluence. Re-arranging Eq.1.3) where Eth is the pulse energy corresponding to ablation threshold fluence.
A semi-logarithmic plot of D2 versus ln(F0 ) is created and the ablation threshold fluence is found by extrapolating the linear curve resulting from this plot to zero diameter where Fth = F0 . termed incubation . is 24 . it is common to speak of damage threshold fluence Fd above which visible damage (but not removal) of the surface results. (3.3) in Eq.5) Thus. For at least some materials. If N > 1 laser pulses of the same energy are incident on the same surface location.4) yields 2 D 2 = 2 w0 ln(E0 / Eth ) . the ablation threshold fluence for large N is only marginally larger the damage threshold fluence. Using Eq. (3.maintained at a known. it has been found that ablation occurs for pulse energy less than that corresponding to the ablation threshold fluence for N = 1 and that larger N results in more decrease. experimental damage diameters can be measured and Fd can be found by a similar procedure as for Fth . The decrease of the ablation threshold for multiple pulses. which does not vary with the number of pulses .1. Since calculated focus spot sizes are subject to inaccuracy.1. the effective radius of the laser focus spot needed for fluence calculations can be calculated from the slope of a plot of experimental ablation diameter-squared and pulse energy data. At fluences lower than the ablation threshold. If Fd < F0 < Fth . fixed value. (3.2) and Eq.1. (3.1. an effective focus spot radius for experiment can be calculated directly from the ablation data .
attributed to increased absorption due to accumulation of damage or effects (i.1. The affect of incubation on ablation threshold can be quantified by a relationship of the form  Fth ( N ) = Fth (1) N ξ −1 (3. focused to a Gaussian spot radius w0 .6) where N is number of pulses incident on a given location and ξ is called the incubation factor.1.e. color center formation) from individual pulses. scanned at a linear velocity s. When ablation is carried out with a pulsed laser having pulse repetition frequency f. denoted Od.1. it is customary to refer to the ratio of the length along the scan centerline of the intersection of focus spot areas of successive pulses to the Gaussian focus spot diameter as the focus spot overlap.7) 25 . Od can be calculated as ⎛ s ⎞ Od = ⎜ 1 − ⎟ ⎜ df ⎟ ⎝ ⎠ (3. As shown in the geometry illustrated in Fig. 3.
In this paper. pulse energy) smaller than that required for ablation with a single pulse. the diameter of a spot ablated by a single pulse for the laser focus spot diameter. d = 2w0 . The high repetition frequency implies that. calculated by the same formula but substituting D. a condition where ablation may occur at a pulse fluence (and hence. a given location along the scan path can be irradiated with a large number of laser pulses. one may also speak of the overlap of ablated spots. For scanned ablation. we first study single-pulse ablation of ITO 26 . even at relatively high scanning speeds.15] presents interesting possibilities for direct-write thin film patterning by ablation of linear features with a single scanned focused laser spot.1 Geometry of overlapping ablation spots showing calculation of theoretical ablation line. The recent development of high repetition frequency amplified femtosecond lasers [14.Figure 3.
ITO ablation experiments were performed at atmospheric conditions using the beam from a Ti:Al2O3 regenerative amplifier operating at a wavelength of 775 nm. an M2 quality parameter of 1. Pulse energies ranged from 60 nJ to 930 nJ for the experiment.2 Experiments The material samples used in all of the experiments consisted of ITO film with a nominal thickness of 150 nm on 2. The beam had a diameter of 5 mm. The linear optical polarization axis was normal to the scan direction.4 mm achromatic lens (NA = 0. The lower repetition frequency made study of single-pulse ablation convenient with this laser.3 and was focused with a 25.1. The results are then used to analyze and compare the ablation of lines in ITO using femtosecond lasers with pulse repetition rates of 2 kHz and 100 kHz. Output power was attenuated with thin-film polarizing beam splitters to pulse energies of 50 nJ to 3 μJ. Other experiments used an Yb fiber-based chirped pulse amplified laser (IMRA America FCPA μJewel D-400) operating at a wavelength of 1045 nm. line ablation results are extrapolated to predict what possible processing rates with lasers having higher pulse repetition frequency. Finally. pulse duration of 150 fs and a repetition frequency of 2 kHz (Clark-MXR CPA 2110). pulse duration less than 500 fs and a higher repetition frequency of 100 kHz.and glass substrate materials. 3.8 mm-thick soda-lime glass substrate. an M2 quality 27 .25). The beam had a diameter of 5 mm.10) and a 10x microscope objective (NA = 0.
Veeco) as well as scanning electron and optical microscope images.1 Single pulse ablation diameter and depth Fig. Processing scan speeds ranged from 1.parameter of <1.28) infinity-corrected objective. These values are used in the discussion of line ablation results obtained with both lasers. The beam polarization was converted from linear to circular with a quarter-wave plate.1.3 Results and Discussion Single-pulse and line ablation results for the 2 kHz laser and line ablation results for the 100 kHz laser are discussed separately below. 3.5 to 200 mm/s and single spot ablation performance was assessed from study of non-intersecting ablation spots produced with the lower pulse repetition frequency laser. The single-pulse ablation results quantify the difference in ablation thresholds of ITO film and glass substrate and the depth of film removed by a single pulses of given fluence. The depth and width of ablated spots and lines were measured from profiles obtained with an Atomic Force Microscope (AFM. 3. 3.5 and was focused with a 20x (NA = 0. Quesant Q-Scope 250) and with a stylus profilometer (Dektak IIA.1.2 shows the single spot ablation diameters produced by the Ti:Al2O3 laser with different pulse energies and focus optics. Larger spots were produced by defocusing the beam.3. 28 .
0 60.3 μm for the 0.51 y = 56.21Ln(x) + 232.0 0.4 2 =10.00E-05 Energy (J) Figure 3.0 1.0 Diameter 2 (μ m2) 80.0 20.120.7 J/cm2 for the 0.0 100. The effective laser focus spot diameters calculated from the slope of these curves were 14. 29 .54Ln(x) + 830. 3.3) and 0.2 2 = 5.1 NA 0.25 NA optic and 56.25 NA optic (Fig.1 NA optic.00E-07 1.6 μm for the 0. 2 Single pulse ITO ablation energy-diameter curves produced using different focusing optics with fitted logarithmic curves.0 40.95 J/cm2 for the smaller laser focus of the 0.1 NA 0.25 1.1 NA optic.00E-06 1.00E-08 y = 14. The ablation fluence was calculated using these effective laser focus spot diameters and fluence vs. ablation diameter-squared plots were then extrapolated to zero ablation spot diameter to obtain ITO ablation thresholds of 0.
6234 10.00 15.00 0.00 30.0 Fluence (J/cm2) Figure 3.00 5. fluence for 0. systematic study will be needed to clarify this dependence of ablation threshold on irradiation area.00 35.40. 30 .00 20.00 1. Fth has to be increased to compensate for larger net energy dissipated away from the beam center by the hot electron diffusion.3 ITO squared ablation diameter vs.964Ln(x) + 2. One investigator  has hypothesized that as the beam waist decreases.7 J/cm2.00 25.25 NA optic showing curve fit parameters that correspond to Fth = 0. A subsequent focused.00 10. It is noted that other investigators have also found variations in ablation threshold fluence with focus spot size.0 Ablation Diameter (μ m ) 2 2 y = 15.
4 shows the ablation depth with respect to laser fluence with the 0.0 2.1 NA Figure 3.5 J/cm2.10 NA optic. With both lenses.1 NA focus lenses.2 J/cm2 completely removed the 150 nm layer of ITO and the ablation depth was relatively constant as fluence was further increased. 3. fluence above about 1.0 1.25 NA and 0. For the 0.4 Ablation depth of ITO using different focus optics Fig.200 Ablation depth (nm) 160 120 80 40 0 0.0 4. 31 .25 NA 0. substrate ablation increased ablation depth for fluence above 2.0 Laser fluence (J/cm2) 0.0 3.
Figure 3. peak fluence F0 = 2.5 Surface profile of laser ablation of ITO (2kHz. scanning speed s = 20 mm/s) Fig.5 shows a typical profile of the ITO surface after laser ablation at a scanning speed that produced approximately 25% pulse overlap.5 J/cm2 . 0. There was no evidence of substrate ablation at the spot overlap area and the lines produced electrically-separated ITO regions. Ridges of re-deposited ITO material were located along the edges of the ablated line. The re-deposited material height was between 20 nm and 100 nm relative to 32 . 3.1 NA lens.
3.6 were produced with the 0.the un-ablated ITO surface and minor re-deposited material was noted further from the ablated line.58 12. The single-pulse ablation results shown in Fig.62 ) = 2. The ablation characteristics of the glass substrate are also important for this research.0 J/cm2. 33 . This provides a comfortable margin for removing ITO in single pulses without ablating the glass substrate. 3. The glass ablation threshold is calculated as exp(8.4 because ITO and glass were both ablated by a single pulse in those experiments.1 NA lens. It is smaller than the fluence at which glass ablation was observed in Fig.
6 Single-pulse ablation threshold calculation of glass substrate material showing an ablation threshold of 2.1 NA optic (Ep = 0. three-pulse ablation implies at least 67% pulse overlap would be required to ensure that three pulses 34 .7 J/cm2.00 4.4 J/cm2 and with the 0.00 Ablation diameter2 (um2) 12.617Ln(x) .3 μJ). 3. approximately 3 pulses could completely remove the ITO without a possibility of damaging the glass substrate. the single pulse ablation results illustrate that lines can be ablated in ITO by a multiple-pulse process or a single-pulse process.00 8. Summarizing.00 6.8. For example. well above the ITO ablation threshold of 0.4 at a laser fluence of 0. the ablation depth was approximately 40nm. Thus. For ablating lines.0 y = 12.00 2.0 Fluence (J/cm2) 10.14.5827 Figure 3.0 J/cm2.00 10.00 1. from Fig.00 0.
7 2 kHz multipulse ablation with 0. Figure 3.25 NA lens (F = 0. 3. a single pulse will ensure that ablation depth is 150 nm but the laser pulse energy must be at least 0. Fig.8 μJ for this processing approach.7 shows an example of multi-pulse ablation at low energy fluence (the slightly non-linear ablation pattern corresponds to motion system stepper motor resolution).1 J/cm2 or higher. At a fluence of 1. s = 1.are incident on all locations along the scan centerline.5 J/cm2.55 mm/s) 35 .
3. Pulse energies of 80 nJ and above produced lines having electrical separation at the scan speed of 100 mm/s while 250 nJ was required to obtain electrical separation at 200 mm/s.8.2 Line Ablation Results Examples of processing results obtained with the 100 kHz fiber-based laser at a variety of pulse energies and scanning speeds with focus spot size 2w0 = d0 = 3.6 μm are shown in Fig.3. ~250 nJ (bottom) 36 . The results show both complete and partial ablation of lines in the ITO with larger ablated line widths at higher pulse energies.6 μm. scanning speed s = 100 mm/s (top) and 200 mm/s (bottom) and varying Ep showing minimum pulse energy for electrical separation: ~80 nJ (top). Figure 3.8 100 kHz laser ablation results for focus spot diameter d0 = 3.1. 3.
3. The 100 kHz line ablation results were analyzed to determine the ablation threshold fluence for difference spot sizes and scanning speeds with the fiber-based laser.9 Ablation line width vs. 37 . The calculations were done as for single-pulse thresholds but with ablated line width used instead of ablated spot diameter. fluence for f = 100 kHz.The edges of lines ablated with smaller spot size and higher scanning speed were more irregular than those obtained with the larger spot size and lower speed and some ablation of the glass substrate is noted at the higher pulse energy (310 nJ) at 100 mm/s. The data corresponding to a sample calculation for laser focus spot diameter d0 = 3. s = 50 mm/s.1 Ablation width 2 (μm2) y = 6. d0 = 3.7366 1 Fluence (J/cm2) 10 Figure 3.6 μm and scanning speed s = 50 mm/s is shown in Fig. equal to the fluence where glass ablation was observed in single pulse ablation tests.9. The data closely follows the expected logarithmic trend. 18 16 14 12 10 8 6 4 2 0 0. The corresponding fluence is 3.0 J/cm2.6 μm.5001Ln(x) + 6.
6 μm and s = 50 mm/s is 2 D 2 = 2w0 ln (F0 Fth ) = 6.63 J/cm2 depending on pulse overlap. 3.5) = 0.1.8) The results of all tests with varying laser focus spot size. The threshold ablation fluence values range from 0.The corresponding ablation threshold is Fth = exp(− 6. Then. an example line width-fluence relationship for d0 = 3.23 J/cm2 to 0.10. pulse energy and pulse overlap are displayed in Fig.35) . This variation of ablation threshold can be explained by the affects of incubation.48 ln (F0 0.35 J/cm2.74 / 6. 38 . (3.
25 J/cm2 reported elsewhere .7 J/cm2 obtained using the Ti:Al2O3 laser whereas at high pulse overlap.1 0 40% 50% 60% 70% 80% 90% 100% 6. The affect of incubation on ablation threshold data shown Fig.6 0.4 0. When a pulsed laser focus spot is scanned linearly.3 0.2 0.3 um 3.Threshold Fluence (J/cm2) 0. the ablation thresholds approach the ITO damage threshold fluence of ~0. multiple pulses may impinge 39 .10 Variation of line ablation threshold with focus spot overlap at various laser focus spot diameters showing decrease in threshold at high overlap due to incubation.6 um 11 um Focus Spot Overlap (%) Figure 3.5 0. modified to account for the linear scanning of the beam.10 was analyzed using the relations summarized above.7 0. 3. the ablation threshold approaches the single pulse ablation threshold fluence of 0. At low pulse overlap.
Also. it is appropriate to use whole numbers rather than integers in this calculation.1. 3. the number of pulses incident on points laterally displaced from the scan centerline decreases as the cosine of the lateral distance. Again referring to the Fig.1. 40 . (3. falling to a single pulse as distance approaches the beam radius. the number of pulses incident on a single point along the centerline of a linear scan can be approximated as N= d s f = 1 .9) were used to calculate the pulse ablation thresholds and incubation factors corresponding to the high frequency scanned laser ablation results of Fig.on points along the scan path so incubation may be used to analyze the scanned ablation results. 1 − Od (3. Pulse numbers calculated from Eq. 3.10. the results are displayed in Table 3. A convenient measure of this affect is a computation of the number of pulses N incident on a single point along the scan line.9) Since the beam energy extends beyond the Gaussian radius assumed in the calculations.1.1.
71J/cm2to 0.7 J/cm2 and 0.10 and the results for ablated line diameter versus fluence allow the maximum ITO ablation speeds for higher repetition frequency lasers to be assessed with the affects of incubation included.3. The calculated single ablation thresholds range from 0. which were 0. increasing line quality. The data displayed in Fig. which makes the ablated lines wider and also increases the overlap of ablated spots.1 Single pulse ablation thresholds and incubation factors for lines ablated at with different focus spot sizes.Focus spot diameter.6 6. A straightforward 41 .7745 Table 3. incubation lowers the ablation threshold at higher pulse overlap. ξ 0. d0 μm 3.82 Incubation factor.76 0.82J/cm2 and are comparable to those obtained for single pulse ablation with the Ti:Al2O3 laser.71 0. The thresholds and incubation factors obtained at different spot sizes are slightly different.95 J/cm2 . as was also the case with the single pulse experiments.6469 0.779 0.3 11 Calculated single pulse ablation threshold. For fixed pulse energy. Fth(1) J/cm2 0.
pulse repetition frequency f and focus spot size w0 corresponding to a laser and focusing NA of the optic of interest and a maximum ablation line width. Δ D . are chosen as application requirements. by the geometry of overlapping ablation spots illustrated in Fig. the data corresponding to d0 = 3. and fractional line width variation. 3. Then. D. It is noted that. 3. 1 − (1 − Δ )2 = s Df .11. For example. corresponding to a focus optic of interest.calculation allows prediction of the ablation performance of higher repetition frequency lasers. one begins with data from one curve of Fig.1.10) 42 . Briefly.6 μm may be used.10. (3.
4) cannot be produced by the laser or is less than the ablation threshold.7) and a corresponding ablation threshold for ITO is found from the focus spot overlap vs.Δ/2 D/2 D ⎛ s ⎞ 1− ⎜ ⎟ 2 ⎝ Df ⎠ 2 s/f Figure 3. 3. ablation threshold relationship from data plotted in Fig. (3. (3.6 μm and ablation line width D = 2. Focus spot overlap can be found by Eq.11 Geometry of overlapping ablation spots showing calculation of theoretical ablation line ripple as a function of ablation spot diameter and processing parameters. Then.1. consider f = 100 kHz.1.10) and Od = 76% from Eq.7). s = 86 mm/s from Eq. (3. An ablation 43 . the scanning speed s and all other ablation parameters can be calculated. To illustrate the procedure and the correctness of the calculations.1.1. Thus. (3.75 μm and line width variation of 5 %. If the fluence required by Eq. then the focus spot size must to be adjusted and calculations repeated. d0 = 3.10.
A plot of scanning speed versus pulse repetition frequency for a focus spot diameter of d0 = 3. fluence F0 = 1. 3. (3. an image of a line ablated at s = 100 mm/s . 3. One goal of the analysis is extrapolation of the data obtained from the experiments to estimate ablation performance at higher pulse repetition frequencies.6 μm from such a calculation is shown in Fig.4) and Ep = 78 nJ from Eq. The measured width of 2. 3.6 μm and estimated width variation of 5% are in reasonable correspondence with the predicted values.1.threshold of 0. (3.48 J/cm2 is found from Fig.2).10. 44 .12.8.1. For comparison. Ep = 80 nJ and Od = 72% is shown in Fig.54 J/cm2 from Eq.
For example the line width and fluence of the 65% overlap curve are 2. Ep = 100 nJ and d0 = 3.E+07 Figure 3.3 45 . pulse repetition frequency predicted from the results for pulse repetition frequency f = 100 kHz.00E+00 94% 67% 1. the ablation line widths and fluences for the two spot sizes are quite different.6 μm for ablated spot overlap of 94% and 67%. From this figure.12 Scan speed vs. corresponding to pulse overlaps of 64% to 97%.65 J/cm2 with 6. Similarly.0 to10.00E+01 Scan speed (m/s) 1.0 m/s are predicted.00E-02 1. However. a frequency of 5 MHz would be expected to yield reasonable line quality at scanning speeds from 1.0 m/s to 5. A plot of data from the same analysis for a focus spot diameter of 6.E+06 Pulse repetition rate (Hz) 1. scanning speeds ranging from 2.00E-01 1.6 μm focus spot size.E+05 1. at 10 MHz pulse repetition frequency.3 μm produces the same curve as the 3.5 μm and 0.0 m/s.1.
μm focus spot diameter. which are well-matched to the calculated values.3 μm optic and Ep = 185 nJ is shown in Fig.6 μm focus spot.96 J/cm2 with the 3. 3. The reason for the variation of line width with overlap is variation of ablation threshold due to incubation. Two available experimental data points. 46 .13. To illustrate this affect. are also plotted. the variation of line width with scanning speed for the 6. Line width and fluence of the 67% overlap curve are 3 μm and 1.
the lower ablation threshold due to incubation could be used to advantage if the pulse energy of the available laser were limited. However.2 0.3 Exp Figure 3. For a pulse energy of only 20 nJ.025 m/s has a pulse overlap of 96% and an ablation threshold It is interesting that the highest speed occurs at the lowest pulse overlap.3 μm and Ep = 185 nJ showing the effect of lower ablation threshold at the lower speeds.05 0.1 0.5 5 4.6 Ablation line width ( μm) 5.13 Predicted and experimental ablation width versus scanning speed for d0 = 6.25 Scanning speed (m/s) 6.7 J/cm2.15 0. allowing for 47 . indicating that benefits of lower ablation threshold produced by incubation from highly overlapping pulses do not overcome the speed penalty incurred to high pulse overlap. The experimental point at 0.4 J/cm2.5 0 0.5 4 3. corresponding to a fluence of 0. single-pulse ablation would not be possible with this laser as the threshold fluence is approximately 0. However. 6.3 Calc.
Further. (3.considerable pulse overlap with a 10 MHz laser.  mentioned earlier in this report.18].1. s = f ⋅ d ⋅ (1 − Od ) (3. defining c' as the slope of the scanning speed vs.7) to obtain an expression for scanning speed. (3. pulse repetition rate graph so that s = c ′f and noting that this slope 48 . it is foreseeable that interference from a non-extinguishing plasma above the material surface would eventually set an upper limit on achievable ablation scanning speed [17. the pulse energy does not influence the ablation scanning speed. pulse energy for given ablated spot overlap. Essentially.25 J/cm2.11) Noting that Eq. the lower threshold due to incubation would allow ablation of quality lines at high speeds: 85% pulse overlap at 2. it would not be expected that much further decrease in ablation threshold could be achieved by higher pulse overlap.1. One effect that may limit the maximum pulse repetition frequency for effective laser ablation is related to plasma formation. an alternative equation for scanning speed is s = f ⋅ D ⋅ (1 − OD ) (3.1. which are the same as the damage thresholds reported by Ashkenasi et al. while an applied fluence above the threshold fluence (including incubation) is necessary for ablating a scan line. As pulse repetition frequency is increased.5 m/s in this case. With regard to the lowering of the ablation threshold by incubation.12) where OD is ablation spot diameter overlap.7) also applies when ablated spot overlap is used instead of focus spot overlap. Thus. One result is noticed when performing calculations of scan speed vs.1. it is also noted that the ablation thresholds achieved in this work were on the order of 0. This can be seen by rearranging Eq. once this requirement has been met.
5 J/cm2 could completely remove the ITO without substrate damage. 3. ds = 3. it is clear that scanning speed does not depend on energy once either ablated spot overlap or focus spot overlap has been fixed.1. Analysis of the ablation performance of higher pulse repetition frequency lasers was used to predict performance of higher pulse repetition frequency lasers.equals the distance traveled during one pulse repetition period (the “pitch”). so single pulses with F = 1. They have much potential for use in 49 . Taking advantage of the additive ablation of several overlapping low energy pulses.0 J/cm2 – 2. ablation of lines at 1. Lines were ablated in ITO with the 100 kHz pulse repetition frequency laser. Due to the highly overlapping pulses. The single pulse ablation threshold of ITO was substantially less than glass.4 Conclusions and future works Experimental results for the femtosecond laser ablation of ITO films from glass at different pulse repetition rates were analyzed.5 mm/s was demonstrated with the 2 kHz system at Ep = 300 nJ. Ablation of good quality lines at the s = 100 mm/s. Recently. The feasibility study showed that the femtosecond laser can be used for ITO ablation. lasers with higher pulse rates (> 1MHz) and sufficient pulse energy for micromachining have been developed.6 μm and Ep = 80 nJ was demonstrated. femtosecond laser pulse ablation was effective for patterning ITO film at relatively lower pulse energies and higher scanning speeds. but low productivity or low scanning speed can be still an issue for production.
The scanner application needs more study in future work for assessment of scanning mirrors damage thresholds and focus spot diameter limitations.micromachining applications. 50 . The motion system may be the limitation for machining with these high pulse rate systems. but high speed scanners would be expected to solve this issue.
Kim. Fan. Ocana. Rev. 43. J. N.H. 85. 2631-2645 (1998). Schulzgen.1. Giridhar. Mansuripur. Elbandrawy and M. Opt.” Appl. Opt. Appl. C. J. M. C. 3124-3131 (2001). Opt. Khulbe. “Femtosecond pulsed laser micromachining of glass substrates with application to microfluidic devices.J. O.” Appl. “Coulomb explosion in ultrashort pulsed laser ablation of Al2O3. Urbina and J.” Synthetic Metals 138. “Effect of substrate absorption on the efficiency of laser patterning of indium tin oxide thin films. Rosenfeld and EEB Campbell.13173 (2000). “Indium tin oxide overlayered waveguides for sensor applications. Seong. 7137-7143 (2006). “Modeling optical breakdown in dielectrics during ultrafast laser processing. “Tin doped indium oxide thin films: Electrical properties. 5. L. “Optical characteristics of femtosecond laser micromachined periodic structures in Si <100>. Opt. P. App. 45. C. 6. Carabe. 36. “Microprocessing of ITO and a-Si thin films using ns laser sources.B.T. Microeng. 7066-7072 (1997).H. Micromech.” Appl. 83. R. Kim.N. M. B 62.J. D. Kim. “Novel patterning method using Nd : YAG and Nd : YVO4 lasers for organic light emitting diodes. 4207-4212 (1999) 4.” J. Kautek. J. S. Kim. M.S. 9.” Phys. 13167. Luff. Lee and S. “Ultrashort Pulse Laser Interaction with Dielectrics and Polymers. Krüger and W. Y. J. R. 4584-4589 (2004). 15.P. Peyghambarian and M. 8. Ban. Molpeceres . Gupta.3.” Appl.” Adv Polymer Sci 168. Ashkenasi. Takai. T. A. Gandia.” J. 2. B. Stoian. M. Y. A. Longtin. 10. Lauzurica . K. Yavas and M. 51 . Wilkinson and G. Phys. 40.S.5 References 1. Tak.” J. 7. K. Bel Hadj Tahar. Ohya and Yasutaka Takahashi.L. 3. J. Perrone. 247–289 (2004). 497-500 (2003).H. Phys. 1271-1278 (2005).
11. A.A. M. Fisette and M. Ashkenasi.” Opt. Wise. Lorenz. 223-228 (2003). D. Bulgakova and E. Dausinger. “High energy femtosecond Yb cubicon fiber amplifier. G. J. 17. Klimentov.” Appl. P. Prokhorov. I. 96. Sci. Shah.O.E. 150. Byer.” Opt. Cho and M. Hertel. Sci. Fermann. 13. Salle. Muller. 15. “Surface damage threshold and structuring of dielectrics using femtosecond laser pulses: the role of incubation. A. Lett. Elec. 1. Meunier. L. Imeshev.C. A-Mater. R. 77. G. “Fundamentals and advantages of ultrafast micro-structuring of transparent materials. V. Buckley. Gobert. Stoian and A. Semerok. Processing. 52 . S. Pivovarov.V. B. Sosnowski. Z. Ben-Yakar and R.R. Phys. 101-106 (1999). Breitling and F. 1888-1890 (2005).M. Phys. “Femtosecond fiber lasers with pulse energies above 10 nJ. 18. M. P.V. G. B. S. N.” JLMN-J. G. Express 13. 12. Rosenfeld. S381-S383 (1999). Ashkenasi.B. “Femtosecond laser ablation properties of borosilicate glass.M.” J. Processing 69. 378-382 (2001). Stoian. D. 5316-5323 (2004). 31.” Quant. Garnov. Phys. Femtosecond and picosecond laser microablation: ablation efficiency and laser microplasma expansion. Meynadier. 16. R. Ilday and T. F.” Appl. T. Appl. Petite and A. Hartl . Surface Sci. Konov. Perdrix.V.M.” Appl. O. Laser Micro/Nanoengineering.E.Rosenfeld. 14.L. Kononenko. I. F. Liu. A – Mat. A. 4717-4722 (2005). “Three-dimensional microfabrication inside photosensitive glasses by femtosecond laser. “The role of plasma in ablation of materials by ultrashort laser pulses.L. Campbell.I. 7-11 (2006). 30. D.W.
4] and direct-write laser micromachining [5-7] have been used to fabricate prototype or small lots of polymer microfluidic devices. The hot embossing method has been used to fabricate inexpensive. precise polymer microstructures for optical components or microfluidic devices [2. hot embossing  is more commonly used for the finest features. polymer and metals. Relative to the surface micromachining 53 . Some surface micromachining processes. economical polymer manufacturing. There are a number of options for fabrication of microfluid channels in polymer. Of the mass production processes that are capable of high-rate manufacturing of polymer microfluidic products. However. low conductivity for electrokinetic pumping or electrophoretic separation.3.and nanoscale fluid flow channels in polymer materials are needed for fabrication of many biomedical devices. The tool can be made from a variety of materials including silicon.2 Hot embossing 3. 8-10].1 Introduction Micro. Polymer materials are advantageous for such devices because they offer a wide range of material properties including good biocompatibility . low material cost for high-volume mass fabrication of disposable devices and replication methods for mass production .2. notably plasma etching [3. mass production techniques such as embossing or injection molding are best suited for large scale. a tool impresses a pattern into the surface of a polymer that is softened by heating above its glass transition temperature. In hot embossing.
However. the choice of which depending on the material. Laser and electron beam lithography and focused ion beam (FIB) milling techniques have high resolution for generating microscale features as well as nanoscale ones in durable metal tool materials. Also. which is not as problematical with the more durable metal mold materials . CNC machining is used for creating larger features in a wide range of metals. FIB has low throughput even considering that the manufacturing cost of the tool will be used to manufacture many polymer parts. LIGA. a process that combines E-Beam lithography and electroplating to create metal parts with microscale features has also been applied for embossing tool fabrication but E-Beam lithography apparatus is expensive. precision and productivity requirements. hot embossing offers much higher productivity. lithographic patterning requires a costly clean-room environment and needs potentially-costly multiple fabrication steps [12. A disadvantage of the silicon tools is a tendency to break during demolding. but is not useful for submicron feature sizes. Hot embossing tools may be created with a variety of processes. For silicon and metal films. It is well known fact  that the femtosecond lasers with typical pulse durations of about 150fs can ablate material with significantly lower thermal effect on the residual 54 . many micromachining processes commonly used in electronics fabrication may be applied for creating microscale and nanoscale features. 13]. comparable feature sizes and lower cost for forming precision shapes in polymer.processes mentioned previously. The hot embossing process is compatible with a variety of polymers with a wide range of material properties and surface chemistries for specific applications.
material than nanosecond lasers with pulse width ranges from 10 to 100’s of nanoseconds. and ceramics . However. However. Studies have shown that the heat affect zone (HAZ) width with femtosecond laser can be a few hundred times lower than for nanosecond pulses . It is capable of making microstructures with feature sizes on the order of 1 μm and applicable for polymers. excimer laser ablation still has some issues of heat input and is more favorable for materials with low thermal conductivity and low melting temperature. Many metals and carbides which would make very rugged. such as glass. A series of experimental investigations were carried out to analyze the effect of ablation process parameters on achievable feature size of 50μm (minimum 8μm). and material removal rate 55 .and nanoscale features without material limitations . surface roughness of 285nm (RMS). free standing trench angle of 36 degrees. An alternative laser used for direct-write ablation of metals is the excimer laser. the same superior mechanical properties also create difficulties for mechanical machining. metals. stress-induced martensitic transformation at the cutting front leads to an undesirable contact interface between the tool and the workpiece . The combination of hot embossing of polymer materials with tools flexibly micromachined by femtosecond laser ablation is seen as a potentially advantageous process for mass production of biomedical devices with microscale and sub-microscale features. Also. The objectives of this research were to characterize capabilities of femtosecond laser micromachining for creating embossing tools of stainless steel. femtosecond laser micromaching can make micro. durable embossing molds are not compatible with deposition and micromachining processes often used in microfabrication.
3. This investigation focused on fabrication of hot embossing molds in polished AISI 304L stainless steel blanks by femtosecond laser micromachining.2 Experiment procedure Steps for polymer microfluid device fabrication illustrated in Fig.88 μm3/pulse at 500nJ which can remove material at 5.652 μm3/s with 6kHz pulse repetition rate. 56 . 3. hot embossing and cover bonding.measured to be is 0.14 including mold fabrication.2.
Figure 3. the roughness of the mold material directly affects to the roughness of the microfluid passages. Since the final embossed patterns replicate the features on the mold. The average surface roughness Ra was 12. An image of the optical profilometer results characterizing the polished surface is shown in Fig.15.21mm x 0.37 nm and the root mean square 57 . The surface profile was measured using an optical profilometer (Veeco Wyko DMEMS NT3300) with an array size of 736 by 480 points and resolution of 1. Thus.3. long-life hotembossing tools . the surface of the mold blank was prepared by precision polishing (Ultra Tec MULTIPOL 8) using 3μm aluminum oxide.92mm. (b) femtosecond laser machining of mold. The surface height data was processed to remove the tilt of the planar surface.14 Steps for fabrication of a polymer microfluid device by the hot embossing process: (a) precision polishing of mold blank. (c) hot embossing.65μm. covering an area 1. (d) assembly of device with top cover. The mold material was selected as one typically used for durable.
surface roughness Rq was 16. Figure 3.80nm. This value is denoted Rz. A more reliable value of the maximum variation of surface height can be found as the difference of the averages of the 10 highest profile points and the 10 lowest profile points.43nm. and its value was 233.10nm. The difference between the maximum and minimum surface heights in the measuring area Rt was 355. 58 .15 Surface profile of polished mold blank as measured by optical interferometry.
A coaxial vision system was integrated into the laser beam delivery optics to provide a capability for accurate focusing. The polarization angle of the linearly polarized beam emitted by the laser is horizontal (parallel to the optical table) and is rotated by the λ/2 wave plate. The irradiance distribution of the laser output beam is nominally Gaussian. The amount of power transmitted by the second PBS varies depending on the angle of the polarization axis of the incoming light with respect to the optical axes of the beam splitter . The focus location was adjusted by the motorized z axis and a manual micrometer by visual inspection of the coaxial image 59 .6W.16 Since the typical range of power for directwrite micromachining is less than 5mW. The 1/e2 beam diameter and quality factor were measured to be 5mm and M2 = 1. adjustable pulse repetition rate fp = 2kHz and 3kHz. a detailed optical schematic of the laser micromachining system is shown in Fig. 3. Laser power was switched by a high speed mechanical shutter with 4ms control time (NMlaser LS055) in coordination with the part motion. measured full width half maximum (FWHM) pulse width Tp = 150fs and average output power of 2. the laser output power was attenuated by a combination of two thin-film polarizing beam splitters (PBS) and a λ/2 wave plate polarizing rotation optics. The laser has a central wavelength λ = 775nm. The laser beam was delivered to the mold surface for the machining process by beam delivery optics.42 (Mitutoyo M Plan Apo NIR 50x).3.The polished mold blanks were patterned with a laser micromachining system comprised of a Ti:Al2O3 femtosecond laser (Clark-MXR CPA2110) and a submicron resolution motion system. The attenuated laser beam was focused onto the mold surface by an infinity corrected long working distance microscope objective lens with numerical aperture NA = 0.
comparable to the 1.16 Block diagram of beam delivery system setup 60 .generated by a CCD camera-based vision system. Figure 3.6μm focal depth of the 50x objective.1μm. It was estimated that the vision system allowed laser beam focus to be positioned relative to the material surface with an accuracy of +/.
5μm. and hence the maximum fluence. occurs at the focus waist. it is helpful to determine the material ablation threshold.2. defined as the minimum laser energy density that can ablate the material. The axes were driven by linear motors controlled by a computer numerical control (CNC) system (Parker Compumotor ACR9000) and their positioning accuracy was measured by optical encoders with resolution of 0. 3. For the multimode laser beam generated by the laser used in these experiments.2. The relations used in the analysis of laser ablation are briefly summarized and the results of experiments that were conducted to determine ablation threshold and suitable laser machining process variables are discussed below. The focusing optic was manipulated in the Z coordinate axis by a linear stage with a motion range of 25mm.3 Laser ablation To determine appropriate laser micromachining process variables.4 Ablation and damage thresholds and pattern resolution The minimum diameter of a focused laser beam. 3. the diffraction-limited minimum focus spot sized d may be estimated as 61 .The mold blank was fixtured on an XY stage (Parker Daedl MX80) with motion range of 50mm.
d= 2 π ⋅M2⋅ λ NA (3. The radius at which the laser beam fluence equals the ablation threshold fluence Fth can be found by solving the Gaussian intensity distribution equation 2 Fth = F0 exp − 2r 2 w0 ( ) (3. (3.2.1) where NA is numerical aperture.2) where w0 is Gaussian beam radius (half the Gaussian beam diameter) and F0 is maximum (or peak) fluence.3). the calculated minimum focus spot diameter is d = 1. M2 is beam quality factor and λ is the wavelength of the laser radiation. (3. However. as will be discussed below. when performing experiments to measure the material ablation threshold. it is more accurate to calculate the focus spot diameter from the experimental data.18.104.22.168) From Eq. related to pulse energy Ep by 2 F0 = 2 E p / πw0 . For the laser beam and focusing optic used in this work. one may also write 62 .6μm.
2.2.4) in Eq. The effective focus spot radius corresponding to ablation experiments can be calculated directly from the ablation data . (3.2 Fth = 2 Eth / πw0 . (3.2. laser pulse energy by 63 .4) where Eth is the pulse energy corresponding to ablation threshold fluence.2) to solve for the diameter D of the area near the center of the beam from which material is ablated results in the relation 2 2 D 2 = 2w0 ln(F0 / Fth ) = 2w0 [ln (F0 ) − ln (Fth )] (3.6) Thus. (3.2. fixed value. the effective radius of the laser focus spot needed for subsequent calculations can be computed from the slope m of a semi-logarithmic plot of D2 vs. (3. Pulse energy was varied over a range of values while focus spot radius was maintained at a known. Re-arranging Eq.2. Using Eq.5) yields 2 D 2 = 2 w0 ln(E0 / Eth ) (3.3) and Eq.2.2. (3. A semilogarithmic plot of D2 versus ln (F0 ) was created and the ablation threshold fluence was found by extrapolating the linear curve resulting from this plot to zero diameter where Fth = F0 .5) Single pulse ablation spots were produced by scanning the focus spot over the flat surface at a sufficient speed that the spots were separated.
To determine ablation threshold.18.2. 3.6) and the resulting plot of squared ablation spot diameter versus pulse energy and laser fluence are shown in Fig.7) At fluences lower than the ablation threshold. The ablation patterns were imaged by scanning electron microscope (SEM) at an acceleration voltage of 20kV. 64 . (3. The outer diameter of each ablation spot was measured and squared for use in Eq. 3. If Fd < F0 < Fth. 2 (3. experimental damage diameters can be measured and Fd can be found by a similar procedure as for Fth. Examples of typical ablation spot images for each of the powers used in the experiment are presented in Fig.ω0 = m . The average diameter from all spots in a scan row was calculated. it is common to speak of damage threshold fluence Fd above which visible damage (but not removal) of the surface results.2. linear ablation scans were conducted with laser energy set to values in the range from 20nJ to 750nJ and a scanning speed of 30mm/s. selected to ensure separation of adjacent ablation spots.17.
17 Single pulse ablation patterns for various pulse energies 65 .Figure 3.
(b) squared ablation spot diameter vs. peak energy. laser fluence 66 .18 Ablation diameter results with regression-fit lines: (a) squared ablation spot diameter vs.(a) (b) Figure 3.
so the corresponding focus spot radius and diameter calculated from the ablation data are w0 = 1.4μm.2μm and d = 2. expulsion of melt and deep drilling at the center of ablation spots were observed at high laser pulse energy ( > 375nJ).02 (3. It is noted that the focus spot diameter calculated from the ablation data is comparable to the diffraction-limited focus spot diameter of 1. consistent with the observations by other investigators . The regression fit line is plotted with the experimental data in Fig.18(a) and its equation is D 2 = 2. A linear equation was regression fit to the semi-logarithmic plot of squared ablation diameter vs. 3. Conversely.2. Such phenomena are characteristic of high energy femtosecond laser ablation of metal where ablated material forms persistent plasma which radiates energy to the processed surface and causes the observed effects . unlike some of the earlier work.8) where D is expressed in μm and Ep in nJ. pulse energy data. the ablation spots were all due to a single laser pulse so there was no opportunity for the ripples to grow by self-induced interference of scattered and incoming laser radiation. 67 . The slope of the linear function is 2.83 Ln(E p ) . The ripple patterns observed for low fluence ablation had a direction identical to the laser polarization and their period is comparable to the laser wavelength. only slight ablation with practically no melting was observed at lower laser energies (< 50nJ).83μm2/nJ.6μm calculated from the laser beam characteristics and focusing optic parameters using Eq.6. The ripple patterns were indistinct because.Extensive melting.
2. It should be also noted that the ablation threshold differs for single pulse and multiple pulse ablation experiments. it is reported that the incubation effect is also significant for metals .1-0.83 Ln(F) + 4.10) .9) The X-intercepts calculated from the linear functions correspond to an ablation threshold energy of Eth = 4. The equation of the regression-fit line shown in Fig.(3.1).19J/cm2.18(b) is D 2 = 2. The calculated ablation threshold is comparable to the value of 0. 3. the laser focus spot overlap was 75% so a maximum of sixteen pulses impinged on any location within a machined area. however. 316L stainless steel . For the process parameters used in most of the machining in this research (described in more detail below).2 J/cm2 .2. Then.71nJ and an ablation threshold of Fth = 0.71 (3. multiple laser pulses impinge on any selected point within the machined area. The significance of incubation was first noted in the context of dielectric material ablation. is Fth ( N ) = Fth (1)N S −1 68 (3.21J/cm2 found in previous research work for a similar material. The latter situation occurs during machining of lines or areas with a scanned focus spot where the scan speed is slow relative to the pulse repetition frequency. The relationship used to predict the ablation threshold at a location with impingement of N pulses. The single-shot ablation threshold is typically higher than the multi-shot threshold due to the incubation effect. denoted Fth ( N ) . Other literature sources show that the ablation threshold of common metals is in the range of 0.2. All fluences in this work were calculated using this spot size calculated from the ablation data.
3. The measured ablation volume is presented in Fig. Using Eq.09 J/cm2. 69 . Due to axisymmetry of the laser beam fluence distribution. The diameters of these damaged regions was estimated from the SEM images and the same curve fitting procedure used to determine the ablation threshold was repeated to determine a damage threshold.10) with the single-pulse ablation threshold calculated above and an incubation coefficient for AISI 316L stainless steel . To study the efficiency of laser ablation. To make this measurement. three-dimensional profiles of ablation spots were acquired with an Atomic Force Microscope (AFM.17. profile scans are compared to SEM pictures of the same ablation spot from Fig 3. Most ablation spots in Fig.20 and a monotonic increase with pulse energy is observed. 3.2.16 J/cm2.17 have clearly visible damaged areas surrounding the ablation crater.19. the volume of single pulse ablation craters was measured. an estimated sixteen-pulse ablation threshold is Fth (16) = 0. the actual depth profile of the ablation cavity was also approximately axisymmetric so the volume of the ablation cavity was estimated by integration based on the AFM profile scan . In Fig. The final result of this calculation was a damage threshold of 0. (3. 3. Digital Instruments Nanoscope).where S is an incubation threshold.
Right E=65nJ) 70 .Figure 3.19 3D profile of laser ablation (Left E=500nJ.
Assuming that femtosecond laser ablation removes all material from the ablation crater by evaporation.Figure 3. A basic assumption inherent in such a calculation is that all material within the ablation volume just attains energy equal to the evaporation enthalpy at the boiling 71 .20 Ablation volume as a function of laser pulse energy The ablation efficiency is a possible criterion for optimizing the laser machining process. one can estimate the efficiency of material removal from the volume data.
21.2. There are several sources of error in such a simple efficiency calculation. A maximum efficiency of 2% is predicted at relatively low pulse energy of 65nJ. 3.temperature. The images in Fig. Ejected molten material could contain less energy than evaporated material since it only needs to be heated above the melting temperature to be ejected. Tv is evaporation temperature (assumed equal to boiling). This is a 72 . where plasma is visible above the ablation area. V is ablation volume. A formula for ablation efficiency η based on this concept can be written as η= ρ [C p * (Tm − T0 ) + Lm + C p * (Tv − Tm ) + Lv ] Ep ∗V (3.11) where C p is average specific heat of the material. (3. It is observed that ablation efficiency is lower at low laser energy. One is due to the likelihood that much of the ablated material has considerably more energy (thermal and kinetic) than that assumed.11) is plotted as a function of pulse energy in Fig. T0 is room temperature. This is clearly the case at high pulse energy.17 confirm that melt ejection was significant for experiments at the medium and high range of laser energies. The efficiency then is computed as the ratio of this energy to the laser pulse energy. T is melting temperature.2. Ep is pulse energy and Lm and Lv are latent heats for melting and evaporation phase changes. Conversely. 3. The efficiency estimated by Eq. ρ is average density of material. to the extent that melted material is ejected by vapor pressure.
50% 2.21 Ablation efficiency 73 .00% 0 100 200 300 400 500 600 Energy (nJ) Figure 3. 2.17.00% 0.reasonable result since at lower pulse energy. the fluence of most of the laser focus spot area is below the ablation threshold. Thus.50% 1. This is likely because the energy of the ablated material is significantly above the evaporation enthalpy.00% Efficiency 1. The peak ablation efficiency region was found to be in the range from 35nJ ~ 165nJ. the ablation patterns produced at these energies were relatively distinct with minimal recast material and surface damage around the rim of the ablation crater. surface damage rather than ablation is the result over much of the focus spot area. 3. Lower ablation efficiency was also found at the higher energies. From Fig.50% 0.
For each scan pass. the length of intersection of individual ablation spots in the scan direction (the ablation spot “over-lap”) was adjusted to produce uniform material removal. The overlap between individual ablation spots along a scan line depends on the scanning speed and laser fluence. 3. Because of the latter dependence.3. laser focus spot diameter d and overlap Ld are shown. Also. Consecutive scan passes were also over-lapped in the lateral direction by a distance equal to the overlap of ablation spots within the scan passes. 74 . machining depths were required to be larger than the single-pulse ablation depth for reasonable pulse energies.2. the geometry of a typical scanned ablation pass is shown in Fig. overlapped curvilinear machining passes.22 where the scan speed s. Thus. To achieve the required machining depth.5 Mold micromachining Mold machining required the removal of material from areas much larger than the laser focus spot diameter. ablation was carried out in multiple layers with each ablation layer consisting of multiple. it is most convenient to specify the machining process in terms of overlap of focus spots rather than ablation spots.
expressed as p= s .12) The length of intersection of adjacent focus spots corresponding to two consecutive pulses is d – p. so the overlap as a percentage of focus spot diameter is 75 .2.22 Overlap of adjacent focus spot locations corresponding to consecutive laser pulses during a scanned ablation pass illustrating a focus spot overlap of approximately 33% The pitch p shown in this sketch is the scan distance covered in one pulse repletion period.Figure 3. f (3.
3.2. Eqs.13) Assuming known focus spot diameter and pulse repetition frequency. width and width variation of the ablated lines.12) and (3. 75%.Ld = d−p × 100% . 1.6 mm/s. and 500 nJ in a series of single line scan ablation tests.2.14) Scan speeds to produce focus spot overlaps of 25%. (3. and 90% were calculated as 5. 50%. Examples of lines ablated at 50% and 75% overlap and pulse energy 165nJ are shown in Fig. ⎣ ⎝ 100⎠⎦ (3. 76 .2. 250 nJ.13) can be combined and the result re-arranged to yield an expression for the scanning speed that corresponds to a desired focus spot overlap: ⎡ ⎛ L ⎞⎤ s = ⎢1 − ⎜ d ⎟⎥ ⋅ d ⋅ f .23.8 mm/s and 0.2.4 mm/s. 330nJ. d (3. SEM images were examined to determine the smoothness. 3.7 mm/s and each speed was used with pulse energies of 165nJ.
the smoothest most consistent area ablation results were obtained at this lowest value.23 Multipulse overlapped scanning (E=165nJ) Of the energies investigated for multi-pass machining.2.20. although the material removal rate was also lowest. the best compromise between channel quality and ablation speed was judged to be produced for the 75% focus spot overlap.15) 77 . From the slope of the data in Fig.0017⋅ Ep (3. 3.Figure 3. Also. the material removal rate is estimated as V = 0.
It is notable that other investigators (previously cited references) have generally found this to be the case for a variety of materials. it was found that providing a cross-flow of argon gas at 39x10-6 m3/s (5 ft3/hr) from a small nozzle helped to avoid this problem. In our experiments. spot size and focus spot overlap of 75% were used for the present application.2.23. Examples of such debris can be seen in the channels depicted in Fig.where V is ablation volume (μm3/pulse) and Ep is laser pulse energy (nJ). 3. Ablation of SiC at a pulse repletion frequency of 50MHz has been demonstrated . We note that the smoothest and most precise femtosecond laser ablation was obtained for relatively low pulse energies in our work. it was noted that the ablated material tended to re-deposit onto the channel surface. This effect also needs to be minimized in order to obtain the best quality machined surface. 78 . the maximum linear scanning speed would 90m/s. Most of the machining needed for mold fabrication required removal of areas of material with width larger than that removed by a single scan pass. If 50MHz and the same pulse energy. A recent trend in ultrafast pulsed laser development has been toward systems with low pulse energy but high pulse repetition rate. As channels were being machined during this investigation. It was found that area ablation done using the energy and overlap mentioned above and adjacent scan passes spaced for a focus spot overlap of 75% produced satisfactory results.
2. 79 . The designed channel width was 75μm and as shown in Fig.24 (b).6 Hot embossing The developed parameters were used to fabricate a prototype microfluid channel mold for a molecular magnetic separation device.24(b). The taper of the channel feature can be observed in the surface profile in Fig. 3. A total of 5 layers of ablation were required to machine mold features tall enough to produce a channel depth of 10μm and the taper of the channel feature resulted in a channel width of only 25μm at the top of the mold feature.24(a) and 3. the width of the channel feature on the mold was the approximately the same dimension at its base. 3.3.
(b) surface profile of mold 80 . Hot embossing mold on stainless steel.24 (a).100μm (a) (b) Figure 3.
81 . 3. the finished device is depicted in Fig.The hot embossing experiment was conducted on custom-made equipment.25. Finally. Before the embossing. both the mold and the PMMA were heated up to 130°C. The PMMA surface adjacent to the channel had RMS roughness of 285nm.75in x 1/16in) was put on top of bottom mold and the embossing mold fabricated by laser ablation was placed on top of the PMMA. with RMS roughness of 20 nm. The embossing mold was then impressed onto the PMMA surface with an embossing force of 25lb. A sample of PMMA (McMaster Carr) (0. The magnetic separation device fluid channel pattern was replicated on PMMA by hot embossing process. Both the force and the temperature were held constant for 5 minutes. However. the mold was removed and the embossed PMMA sample was separated from it.75in x 0. attached with UV curable glue. approximately equal to that of the polished mold surface. The system consisted of an electromechanical microtester (Instron 5848). The profile of this polymer surface had approximately the same roughness as the laser-ablated mold surface and was the minimum obtainable for the chosen laser ablation parameters. The bottom of the molded channel was smooth. which controlled the embossing speed and force. after which the system was cooled down to 80°C. this roughness is not important for the function of the finished microfluid device since it was covered by the top-lid of the device.
25. but the motion system program was modified to produce a thinner channel. 3. 3. 82 . As shown in Fig.26. The laser ablation parameters remained unchanged from those used to produce the mold in Fig. Attempts to produce a mold for a channel width of 6μm resulted in a feature with a somewhat irregular top surface.25 Image of a hot embossed molecular magnetic separator pattern in PMMA Other laser machining experiments were done to determine the minimum channel width that could be machined into a mold.Figure 3. a mold for a minimum channel width of 8μm was achievable. The height of the channel features was set at 12μm.
A common technique to ablate features smaller than the diffraction-limited focus spot size is to use laser energies that generate peak fluence only marginally larger than the ablation threshold. The achievable mold feature sizes are related to the minimum ablation spot size.2. but 83 .Figure 3. The depth of the well and channel features is 12μm. More laser machining parameter development using these low energy is planned. 3. by careful selection of laser energy.4μm. All ablation spots for which power is lower than 38 nJ have diameters smaller than the focus spot diameter. which are determined by the ablation parameters being used. (3. As shown in Fig.26 SEM images of an embossing tool machined to create 8μm wide channel on AISI 304L stainless steel. It is noted that the material removal rate for such low energies estimated from Eq. one can produce ablation spots with diameter smaller than the laser beam spot size. 2.15) is also small.27.
It is also noted that ablation diameter at higher fluence is wider than theoretically predicted value which can be calculated from the ablation threshold value.restricting use of low energy to machining portions of the mold where the finest features are needed would minimize the decreased productivity.27 Theoretical diameter of ablation and damage regions for a pulse fluence distribution with maximum fluence of 0. an ablation threshold of 0.59 J/cm2. As explained before.19 J/cm2 and a damage threshold of 0.09 J/cm2. At lower fluence. outer irradiated area where laser fluence is still higher than that of ablation threshold was wasted to heat the material. Figure 3. the ablation pattern is more 84 .
Based on data from scanned ablation experiments. 3. The ablation quality is again very clean in mid fluence range as shown in Fig. 3.stochastically distributed rather than deterministic behavior. (c) Ep= 500nJ.28. pulse energy of 250nJ with 75% ablation spot overlap was used to fabricate a hot embossing 85 . fp=0.19 J/cm2.7 Conclusions and future works Femtosecond laser micromachining was used to fabricate hot embossing molds for microscale fluid channel fabrication. Single pulse experiments with a broad range of laser pulse energies from 13nJ to 500nJ were conducted to determine the focus spot size and ablation threshold.58 J/cm2 . The threshold ablation fluence for the AISI 304L mold material was calculated to be 0.3 J/cm2. fp =7.28 Ablation profiles for pulses with various peak fluence: (a) Ep= 13nJ. (a) (b) (c)a Figure 3. fp=22 J/cm2. (b) Ep= 165nJ.2.
The characterization of flow properties such as flow rate.mold for molecular magnetic separator. The molecule magnetic particle separator is designed for the particle size around 2μm. An embossing mold with channel width of 75μm was used with the hot embossing process to produce a microfluid device of PMMA polymer.5. The surface roughness vs. However. pressure gradient. flow behavior of internally machined channels is also an interested topic to be explored. yielding a maximum depth/width aspect ratio of 1. Molds for creating channels with a minimum width of to 8μm were achieved. much smaller particles (~ 50nm) will be tested in the future and corresponding channel size needs to be developed. flow profile across the channel and turbulence remains as a future work. 86 . The corresponding ablation depth was measured as 12μm.
G. Lee. H.. Lab on a Chip. 18(3): p. C. Microeng. 11. C. 2002. Chem. and H. Becker. J. J. Plasma etched polymer microelectrochemical systems. Mistry. Femtosecond pulsed laser direct write production of nano. Hung.3.2. Ke. Tan. S104-S111. Appl. 1542 . 121-127. and H. Ray. and B. and V.1550. 130-135. Zhang. Anal. Cher. K. 2005. Bercier. J. K.. J. 210-215. 1997. 2035-2042. Micromech. 4. 10. Lee. Hong. Microeng. Lett.and microfluidic channels. 45: p. 11: p. Farson. 3(2): p. 2004.. Struct. Lagger. 83: p. Enhanced machining of micron-scale features in microchip molding masters by CNC milling.8 References 1. 2006. McDonald. V. Reymond. Girault. UV laser machined polymer substrates for the development of microdiagnostic systems. and U. 567-573. 87 . G. V. Simon. Rosengarten. Microeng. 8. 20-26. H. M. Mecomber. and S. 15: p. Roberts. Limbach. 9. Micromachined pre-focused 1XN flow switches for continuous sample injection. H. Irizarry. Gong. 3. and P. A. 5. 2. Smart Mater. Vollet. C. 14: p. Hot embossing as a method for the fabrication of polymer high aspect ratio structures. J. 7. 945-949. 2001. Sensors and Actuators. and F. Choi. Yalisove. Kapur. Influence of master fabrication techniques on the characteristics of embossed microfluidic channels. Lu. 88(183113): p. Femtosecond laser bulk micromachining of microfluid channels in poly(methylmethacrylate). S. J. J. Polymerization optimization of SU-8 photoresist and its applications in microfluidic systems and MEMS. Esch. Genova. G. 2003. Fabrication of optical waveguides in thermosetting polymers using hot embossing. P. Rossier. 2000. 1-3. and D. Journal of Laser Applications. 2006. Girault. 2(3): p. Choi. Heim. Michel. 6. 2001. B. G. P. J. D. G. Micromech. Yang. Hurd. 69: p. C. Phys. Lab on a Chip. Gobry. L. Plasma etching of thermoset polymer films for microchannels. 11: p. Holland. 145-150. International Journal of Machine Tools & Manufacture. 2006. C. D. A. M. Rossier. Micromech. G. Huang. and L. Carnal. Hwei.
Yu. Johnson. 20. J. Li. 16. Vacuum assisted thermal bonding of plastic capillary electrophoresis microchip imprinted with stainless steel template. Magee. Audouard. Jo. 18-28. and T. Basting. Lett. 1995. Excimer laser: A new tool for precision micromachining. Wu. 19. Gao. 76 . Huang. Austin. Journal of microelectromechanical systems. Fortunier. Pivovarov. Z. Patzel. S. 1999: Wiley. R. 2005. Huot. 275-287. Breitling. Fabrication of 10nm enclosed nanofluidic channels. picosecond and nanosecond laser ablation of solids. Fabrication of high-aspect-ratio microstructures using excimer laser. R. 14: p. Valette. 15.Xie. J. Phys. and D. Cao. Glynn. Kononenko. Coyne. 378-382.81. 2000. Prokhorov. 9(1): p. Y. A. Applied surface science. Motsegood. Z. Chen. 41: p. V. W. 247: p.. Chichkov. L. 827-847. 2004. Quantum Electronics. Fundamentals of photonics. Optics and Lasers in Engineering. Appl. R. 43: p. Applied surface science. A. Experimental investigations machiniability of Ni50. 81(1): p. P. Garnov. B. Tseng. Su. 23. 2005. Chen. and D. 925 . H. D. 17. 2D Calculations of the thermal effects due to femtosecond laser-metal interaction. T.4. 174-176. D. 2004. Tegenfeldt. Lin. 2004. C. G. Three-dimensional Microchannel fabrication in PDMS elastomer. R.936. 18. and Teich. Dausinger. Tunnermann. C. Momma. 1319-1335. 88 . Smart Materials and Structurs. Shirk. and S.6Ti49. and K. O'Connor. Chou. K. P. J. 1996. 109-115. H. E. S. 1998. Journal of Laser Applications. E. 2002. and P. E. Konov. J. N. Wang. 2003. Ma. A(1038): p. 223: p. Zheng. 239-245. 21. Beebe. Mannion. and R. and F. Femtosecond . Saleh. Molian. The effect of damage accumulation behabiour on ablation thresholds and damage morphology in ultrafast laser micro-machining of common metals in air. Y. and Y.12. The role of plasma in ablation of materials by ultrashort laser pulses. 13. and R. Optical and Quantum electronics. Kovacevic. M. 14. H. Endert. Modeling of waterjet guided laser grooving of silicon. Applied Physics A: Materails Science & Processing. 24. Liu. 63(2): p. Journal of Chromatography. Lerberghe. 2001. 22. S. 31(5): p. S297-S301. A review of ultrashort pulsed laser ablation of materials. and Y. Alvensleben. 10(1): p. F. International Journal of Machine Tools & Manufacture. Klimentov. B. 238-242. Chen. and A. 27(12): p. H. Nolte. S. Harzic.
15: p. 89 . Phys. Reider. 2004. M. Farsari. Microeng. S. and Fotakis. 5316 . Femtosecond laser ablation properties of borosilicate glass. 1786-1789.25. 26.. 96: p. Filippidis. J.5323. Byer. Appl. Zoppel. G. Micromech. and R. J. Efficient femtosecond laser micromachining of bulk 3C-SiC. G. Ben-Yakar. A. 2005.
The second part of this chapter is introducing a fabrication of microfluidic channels in soda-lime glass and fused quartz is described with polymer coating. Cells typically grow only 90 . Its characteristics of light scattering in PCL and PET is also described.1 Poly-caprolactone (PCL) machining 4. femtosecond laser interaction with dielectric materials is presented.CHAPTER 4 FEMTOSECOND LASER INTERACTION WITH DIELECTRIC MATERIALS In this chapter.1. the randomly-deposited structure of electrospun fiber does not allow for spatial control of cells necessary to produce complex organs in vivo. The first part is describing the ablation of electron spun poly-caprolactone and polyethylene terephthalate fiber by femtosecond laser. However.1 Introduction Electrospun nanofiber mesh [1-4] is widely used as a tissue engineering scaffold to support cellular in-growth and proliferation for generation of biological tissues [5-9]. 4.
3-1.3 1. Specific polymers properties are listed in table 4. Tg (C) −65 ~ −60 75 Thermal conductivity W/(m K) 0. in this work.1 Physical properties of PCL and PET 91 .3 Bulk density g/cm3 1. PET mainly provides a material comparison to PCL.24 - Solid density g/cm4 1. Direct-write femtosecond laser ablation is a flexible. Polymer Melting temp. PET is thermoplastic resin of the polyester family used biomedically under the trade name Dacron® as a vascular graft. rapid structuring technique that has the potential to alter this via precise control of surface topography to produce specific microenvironments that can influence cellular growth patterns. Temp.15 Heat Capacity kJ/(kg K) 1. It has higher transition temperatures than PCL (e.07* 0.g.2* * PLC/P CL copoly mer Table 4. PCL and PET are commonly used in biomedical applications.on the exterior surface of electrospun fiber following random seeding techniques .4 ES mesh density g/cm3 0. PCL is a FDAapproved biodegradable polyester.14 1.1. sutures and adhesion barriers and is being investigated as a scaffold for tissue engineering. It has a relatively low melting temperature (Tm ≈ 60°C). It has been used in biomedical applications in drug delivery devices. This article describes the development of procedures for the ablation of linear and shaped grooves and cavity arrays that are to be used in subsequent tissue growth studies.13-0. Tm (C) 58 ~ 63 130-137 PCL PET * starch blend Glass trans. Tm ≈ 260°C).
the surface was rendered more hydrophilic (which typically enhances cellular adhesion) by excimer laser ablation texturing .No prior reports have been found regarding laser ablation of ES fiber meshes but laser structuring of PCL films and membranes has been investigated. 13]. The permeability of ultra-thin PCL produced by bi-axial stretching was enhanced by drilling using a femtosecond laser. Thin films of biodegradable polymeric materials. the thermal conduction length corresponding to the femtosecond laser pulse duration is generally small. poly(ε-caprolactone) have been micro-patterned using a Ti-sapphire femtosecond pulsed laser and an ArF excimer UV laser in ambient conditions . Characterization of femtosecond laser ablated PCL by hydrolytic degradation tests showed that the laser irradiation had little affect. so absorbed energy is not significantly diffused. The 775 nm. it is generally considered that nonlinear processes associated with the irradiance generate free electrons which are then responsible for increased absorption and consequent rapid heating. ionization. The photon energy (1. but amplified femtosecond pulses can nonetheless be strongly absorbed and produce clean ablation in many materials. charge separation and removal of the absorbing material [12. Also. When energetic femtosecond laser pulses are incident on dielectric polymers.6 eV) is too small for photochemical ablation . high irradiance pulses typically has minimal affects on the remaining material. It has been shown by a number of investigations that femtosecond pulsed laser ablation has fewer 92 . 150 fs ultrafast pulsed laser was considered promising for electrospun fiber structuring because ablation with the short.
The structure of ES fiber remaining on the walls of femtosecond laser-ablated grooves in PCL and the regularity of machined grooves are compared to those of grooves produced by nanosecond laser ablation. Thus. This paper describes microscale laser structuring of electrospun (ES) polycaprolactone (PCL) and polyethylene terephthalate (PET) meshes. fiber diameters are also on the order of the laser wavelength. 19]. The fundamentals of femtosecond laser ablation have been summarized elsewhere . For our purposes we can consider single pulse removal of a material with threshold ablation fluence Fth by a beam having a Gaussian radial fluence profile 93 . Key issues for the ablation development include preserving electrospun structure post-ablation. In the case of ES fiber meshes. but the sizes and locations are quite random. determining practical limits on lateral resolution and depth of ablated structures and establishing net ablation rates and process productivity. The fluences are compared to estimates of threshold fluence for ablation of the bulk materials.affects on the remaining polymer surface than nanosecond pulsed ablation at comparable wavelengths [14-17]. Electrospun meshes have some similarities to photonic bandgap crystals consisting of periodically-spaced wavelength-sized structural units that have been widely studied in recent years. The fluence required for ablation as well as the affect of fluence and focus spot scanning speed on the dimensions of ablated grooves are determined. optical interactions likely have similarities to that encountered in characterization of human muscle tissue where radiation is highly scattered [18.
related to pulse energy Ep as 2 F0 = 2 E p / πw0 . Clark-MXR CPA2100) were used for this experiment. the diameter D of the area near the center of the beam from which material is removed is then 2 D 2 = 2w0 ln (F0 / Fth ) . 4.3) Fth can be found by curve fitting a semi-logarithmic plot of D2 measured for various F0 and calculating the intercept at zero fluence.2 F (r ) = F0 exp − 2r 2 w0 ( ) (4.1.1. The 94 .1.1) where w0 is Gaussian beam radius and F0 is maximum (or peak) fluence. (4.2) it is noted that F0 = 2 Fav . the pulse duration was Tp = 150 fs and the pulse repetition frequency was fp = 2 kHz. The maximum output power of the laser was Pav = 1.6 W.2 Experimental Apparatus and Procedure Frequency doubled pulses from a mode-locked erbium-doped fiber laser intensified in a Ti:Al2O3 regenerative amplifier laser (CPA. (4. the ablation threshold fluence.1. If F0 > Fth .
the M2 beam quality parameter was measured with a CCD camera and beam characterization software (Spiricon). a Q-switched Nd:YAG laser (TN-50. The beam polarization was adjusted by rotation of this optical element.6 μm. The numerical aperture when the focusing with the 25 mm achromatic lens was NA = 0. The beam diameter of this laser was db = 1. To accurately measure processing power before experiments. The attenuated laser beam was delivered by a beam delivery mirror train through a mechanical shutter and focused on the material. The majority of the attenuation was at the first thin-film polarizing beam splitter. The beam quality was M2 = 1.9 mm and the quality factor was M2 = 15. A 25 mm focal length achromatic lens (NA = 0.1). where only a couple percent of the optical power passes through to a ½ waveplate.output power was adjusted in two stages. maximum average power Pmax = 50 W was also used. a 10x microscope objective (NA=0. a power meter was inserted after the focusing lens.3 in the horizontal X direction. Spectra Physics) with Tp = 120 ns at a pulse repetition frequency fp = 10kHz.25) and a 20x infinitycorrected microscope objective (also NA=0.4 μm and that of the 0.25 with the 5 mm beam diameter) were the focusing optics used in the femtosecond laser experiments.1 NA lens is approximately 6. Prior to performing the experiments.2 in the vertical Y direction and M2 = 1.25 NA objectives is 2. The calculated minimum focus spot diameter for the femtosecond laser with 0. All ablation of linear grooves was done at a scanning speed of 20 mm/s.037 and the estimated diffraction-limited spot diameter was d0 = 271 μm. allowing the additional attenuation at the second polarizing beam splitter to be adjusted. Primarily for purposes of comparison. 95 .
1.4. The tip-to-substrate distance was 16 cm. The ES PET was prepared using 12wt% PET (Mw 18.000. Louis. NJ) and pumped through a 20 gauge stainless steel blunt tip needle at a flow rate of 24 ml/hr under an applied voltage of 24 kV. material flow rate was 18ml/hr and needle tip-substrate distance was 18cm. MO) dissolved in acetone (Mallinckroff Chemicals. St. St. Louis. MO) in a 50/50 solution of trifluoroacetic acid/dichloromethane.000. The mechanical and electrical properties of PCL and PET (such as dielectric values and mechanical viscosity) affect the electrospinning process so the electrospun fiber diameter and fiber density are different.Electrospun PCL was produced using a 12 wt% solution of PCL (Mw 65. Phillipsburg. The density of the resulting mesh and other properties of the polymers are summarized in table 4. Aldrich. The fiber diameter ranged from 0. The distribution of fiber sizes is given in fig.1. Electrospinning voltage was 22kV.2 μm to more than 2 μm. Aldrich. 96 .
97 .6 0.4 1.1 Fiber diameter size distribution of ES-PCL meshes used in the experiments.9 1 1.3 1.3 0.9 8 7 6 5 4 3 2 1 0 0.8 0.2 0.7 1. the directional hemispherical transmittance of different thicknesses of electrospun and solid PCL were measured using an integrating sphere (IS236A. To better understand the penetration of laser and plume energy into the ES PCL.4 0.6 1.1 1.2 1.9 2 More Frequency Frequency Diameter (μm) Figure 4.5 0. NJ). Thorlabs.7 0.8 1. Newton.1 0.5 1. The diameter of the samples used for these experiments was about 20 mm (compared to the unfocused beam diameter of 5 mm).
4.4 μm) while the Q-switched laser pulse overlap was 99.3 % of the larger focus spot diameter and higher pulse frequency. 98 . no ablation was found at less than 3W.4.2 as an initial comparison that established the benefits of the high intensity of pulses for ES nanofiber ablation.3 Results Linear grooves produced by scanning the focused femtosecond and nanosecond laser beams over the surface of electrospun PCL are shown in fig. the femtosecond laser pulses were not overlapped (spot diameter was only 6. The distance between individual pulses on the mesh surface was 10 μm for the femtosecond laser and 2 μm for the Q-switched laser. The higher overlap of Nd:YAG laser with high frequency was caused by the limitation of linear motion system scanning speed. The average power of the femtosecond laser beam was varied between 5 mW and 200 mW while that of the Q-switched laser power was varied between 3W and 10 W.1. Thus. The power of both lasers and the Nd:YAG pulse frequency were varied in a series of experiments to determine feasible values for consistent ablation of microscale grooves.
It was evident from these initial results the femtosecond laser ablated a relatively uniform groove with widths and depths that varied predictably with average power. The width of the femtosecond laser groove is approximately 35 μm. (c) Q-switched laser power was varied from 3 W to 10 W.2 SEM images of (a) Q-switched laser and (b) femtosecond laser grooves. but no power that produced consistent ablation was determined SEM images were used to measure the width of femtosecond-ablated grooves and to characterize the affects of groove ablation on the fibers remaining on the groove walls. Grooves 99 .(a) (b) (c) Figure 4.
4. the high intensity femtosecond laser pulses input energy to the target material over a short time span and diffusion by thermal conduction is commensurately small. which then melts and evaporates depending on the maximum temperature that is achieved. In contrast. in spite of the higher average power. To determine ablation properties electrospun mesh for femtosecond laser irradiation. The energy input by lower intensity Nd:YAG laser pulses is converted to thermal energy in the material. but no significant improvement in consistency of grooves size was found.2(c). grooves were created at different values of fluence by scanning with a focus 100 . practically no ablation was observed as shown in fig. It seems likely that the unstable ablation behavior of the PCL nanofiber for Nd:YAG laser processing is related a transient increase of laser energy absorption associated with fiber melting. The power was varied between 3W and 10W and frequency changed between 5 kHz and 50 kHz. 4. when it did ablate material.produced by the femtosecond laser had minimal melting of the remaining fiber at the edges of the ablation area.2(a). At other times. As mentioned in the introduction. Thus. As shown in fig. the width of the groove produced by the Qswitched laser varied unpredictably from practically 0 mm to almost 1 mm. ionization and ablation tends to occur if the input fluence is over a distinct threshold magnitude. The difference in laser machining results obtained with the Q-switched laser and femtosecond lasers is attributed to the affect of variations in the electrospun mesh density on the different ablation mechanisms for the two lasers. the Q-switched laser induced a high degree of fiber melting on the walls and bottom of the ablation area causing a high loss of the electrospun structure.
The experiments were conducted with 300 μm thick meshes of PCL and PET on a glass substrate coated with 150 nm of transparent conductive film (ITO. 101 . At the same femtosecond laser ablation fluence and scan speed.1 NA focus lens was used for the ablation and scanning speed was 20 mm/s.3 and 4.4 μm. The difference in the material removal rate for the two polymers is at least partly attributed to differences in thermal properties.4. depth and scanning speeds shown in figures 4. The widths of grooves formed are significantly wider than the focus spot size of 6.3 and 4. It is also observed that the width and depth of grooves in PCL are somewhat larger than those formed in PET at the same laser fluence (examples are shown in fig.5 to 3 times smaller.spot for varying laser pulse energy. 4.4 to be in the range of 0.5).1 to 1 mm3/s. the volume removal rate of PET was from 1. 4. The material removal rate for PCL mesh was calculated from the groove width. The groove width and depth were measured for both materials and their variation with pulse energy are plotted in fig. The 0. Indium Tin Oxide) to allow electrospinning.
3 Width of grooves machined in ES nanofiber meshes at various pulse energies using the 0.300 Channel width (μm) 250 200 150 100 50 0 0 10 20 30 Pulse Energy (μJ) 40 ESPCL ESPET Figure 4.4 μm.1 NA optic and speed of 20 mm/s. Error bars indicate error in estimating groove dimension 102 . The groove width is generally much larger than the focus spot size of 6.
103 .1 NA optic and speed of 20 mm/s showing that PCL grooves were slightly deeper.4 Depth of grooves machined in ES nanofiber meshes at various pulse energies using the 0.300 Channel depth (μm) 250 200 150 100 50 0 0 5 10 15 Pulse Energy (μJ) 20 ESPCL ESPET Figure 4. Error bars indicate error is estimating groove dimension.
6. 4. Such a plasma is easily visible at higher fluence conditions and is discussed below. Several instances of thinned fibers and fusion of overlapped fibers are seen in this example. At least some melting similar appears to be difficult to avoid with the femtosecond laser used in these tests.5 Comparison of grooves ablated in (a) PCL and (b) PET Some melting of fibers at the edges of the machined grooves was produced by all of the tests described above. It is obvious that melting and fusing of groove wall fibers is increased with pulse energy and that some melting of the mesh fibers was evident even when the surface was not deeply structured by the ablation. The large grooves produced at high pulse energy may be attributed to melting and ablation of fibers beyond the radius of laser beam impingement by optical radiation from the plasma formed by interaction of laser radiation with fibers. Melting of 104 . A comparison of two grooves ablated in PCL at relatively low pulse energies is shown in fig.(a) (b) Figure 4.
fluence = 9. fluence = 4.1 NA optic at a speed of 20 mm/s and relatively low energy.6 J/cm2). This affect was explored further in experiments discussed below.fibers is also consistent with a conclusion that that femtosecond pulses tended to melt rather than ablate the nanofibers when fluence dropped below the ablation threshold. (a) pulse energy = 1. (b.c) pulse energy = 0.3 J/cm2). cell growth tests currently being carried out indicate that the minimal melting incurred when structuring nanofiber meshes with femtosecond ablation has no detrimental affect on the functionality of the mesh as a tissue scaffold.75 μJ. In any case.6 Two grooves ablated in PCL nanofiber with the 0. The affect of low-fluence femtosecond laser pulses was studied further by varying the distance from the 0.1 NA focus lens to the mesh surface during ablation. Such variations might be expected to occur in practice due to variations in mesh thickness and 105 .5 μJ. (a) (b) (c) Figure 4.
The calculated spot diameter of the defocused beam was 30.7. 106 .7 (a) Focused and (b) defocused (multiple passes. grooves were ablated with a focused beam with power of 5 mW and fluence of 7. As shown in fig. In contrast. For comparison. (a) (b) Figure 4. The corresponding laser fluence was 0. irradiation with the defocused laser beam produced mostly undesirable melting instead of ablation of the fibers. However. scan speed remained at 20 mm/s.7 μm. As with all tests. the higher fluence focused laser beam laser fluence in this test was larger than that the single pass groove was about 65 μm wide and minimal fiber melting was observed on the walls.errors in setting lens position. right) scans showing that the lower fluence due to beam defocusing melted rather than ablated the mesh. 4.3 J/cm2. roughly five times that of the focused beam.7 J/cm2. For producing defocused patterns. the focal point location was set 150 μm above the mesh surface. the laser power was 6mW and multiple scans were spaced 30 μm apart.
The formation of plasma above solid surfaces being machined by femtosecond laser radiation is a well-known limitation on the achievable precision of femtosecond laser ablation in air at ambient conditions and it has been well characterized by other researchers. It has been shown in other thin film ablation work by the authors  that the width of with lines ablated in thin film with femtosecond pulsed lasers still followed the expected logarithmic trend. This procedure is based on an assumption that the groove width and the ablation diameter of a single spot are equal. in accordance with the experimental results. Given the relatively low thermal conductivity. it is believed that heating and melting of fibers by energy radiated from this plasma may be the contribute to wider-than-expected grooves. heat capacity and melting temperature of the electrospun PCL.observed widths of grooves ablated in nanofiber mesh are not consistent with the usual ablation analysis outlined in the introduction where the ablation diameter varies logarithmically with the fluence. the ablation curves for electrospun PCL and PET mesh are non-logarithmic. it is expected that radiation from the plasma would be more effective for increasing the width and depth of grooves in it than in PET. As shown in fig. 4. Based on the observation of a visible plasma during the femtosecond laser machining. The variation of the squared values of ablated groove width with fluence was analyzed to quantify the ablation threshold. The plasma expands rapidly shortly after the femtosecond laser pulse and is observed to persist for many 10’s of nanoseconds [20.8. 107 . 21].
the expected ablation threshold value for bulk PCL  is less than 1 J/cm2. fluence using the 0. so the threshold fluence value of 4 J/cm2 was obtained by extrapolating these points find the x-axis intercept. The PCL ablation fluence threshold suggested by the lower end of the fluence curve is larger than the expected bulk value.5.0 2 Laser fluence (J/cm ) Figure 4.E+04 0.E+04 4. In contrast.8 Groove width squared of nanospun polymer vs.0 100.0 10.1 NA optic showing non-logarithmic behavior.E+04 Width2(μm2) 3. as discussed below.E+04 Fth ~ 4 J/cm2 ES PCL ES PET 1. The slope of the curves is still decreasing at the lower two values plotted for PCL.E+04 2. A detailed analysis of the ablation properties of polymer 108 . The difference is explained when one considers the large surface area of the fibers within a volume defined by the laser focus spot diameter and extinction depth.E+00 1.
in spite of the fact that the measured mesh density was ρm = 0. corresponding to an extinction depth of approximately 120 μm.24 g/cm3.0013 μm-1. 109 . about 20% that of the solid material density. Since it is not a very highly scattering material when solid.nanofiber meshes for femtosecond laser pulses is the subject of current research by the authors. 4.9. the absorption coefficient for the solid PCL was approximated directly from this curve as 0. Thus. To better understand the laser ablation characteristics of electrospun PCL. the extinction coefficient for low-density nanofiber mesh would be expected to be less than that of the solid material. Thus.0085 μm-1. corresponding to an absorption length of about 770 μm. The decrease of mesh transmittance with thickness is well approximated by an exponential function and the extinction coefficient is about 0. the large mesh extinction confirms that scattering rather than absorption is a dominant factor in the attenuation of laser radiation through the mesh. its absorption properties were measured. The results of directional hemispherical transmittance tests of electrospun and solid PCL are plotted in fig. the fiber mesh extinction coefficient is much larger than the solid material absorption coefficient. If extinction were primarily due to absorption.
120% Transmittance (%) 100% 80% 60% 40% 20% 0% 0 100 200 Thickness (μm) 300 Fiber Solid (a) 100% 80% Reflectance (%) 60% 40% 20% 0% 0 100 200 Thickness (μm) 300 (b) Figure 4. 110 . Error bars indicate estimated +/.9 (a) Hemispherical transmittance of ES PCL nanofiber mesh and solid PCL and (b) reflectance of nanofiber mesh for an unfocused femtosecond laser beam.10 μm accuracy in mesh thickness measurement.
such a conclusion needs to be confirmed by measurements of extinction length is a direction parallel to the mesh surface since the mesh appears to be rather nonisotropic in the SEM images. shaped grooves were ablated in the surface of electrospun PCL in preparation for specific tissue growth studies. laser ablation was used to form a 1 mm grid pattern with a microstructured cavity at the center of each grid cell. After the study described above. it is very plausible that the material surface area illuminated by a femtosecond laser pulse is much larger than the focus spot area and thus.The groove depths produced by laser ablation were on the order of the measured fiber mesh extinction length of 120 μm. 4. the ablation threshold determined for fiber mesh is corresponding larger for the fiber mesh than for solid material.10(a). It is also interesting to revisit the ablation threshold estimated for the nanofiber mesh in view of the measured characteristic extinction depth of laser light. Given the that the incoming light is scattered multiple times within the extinction length of 120 nm. A detailed analysis of ablation of nanofiber mesh by femtosecond laser pulses is currently being carried out by the authors. since the lateral penetration of radiation would also be expected to be limited by scattering. the fiber mesh before being absorbed. A typical fiber spacing of 2 μm is estimated from the fiber diameter distribution shown in fig. as would be expected. The groove widths are also of the same order as the extinction length. However. As shown in fig. Ablation parameters were the same as those used to ablate 111 . This is consistent with a hypothesis that radiated plume light is influential in determining the size of ablated features. 4.1 and the fractional density of nanofiber mesh as compared to solid material.
(a) (b) Figure 4.10 (a) Matrix of microscale structures ablated in PCL nanofiber mesh and (b) magnified view of a laser-machined cavity in PCL nanofiber mesh (right) being used in a cell culture test. 4. 112 . a magnified view of a cavity being used for a cell culture test and a cluster of cells can be observed with growth constrained to the bottom of the cavity.10(b). 4.the straight groove in fig. In fig.6.
Preliminary results by a computer simulation of electromagnetic wave propagation inside the PCL fiber showed the wide scattering or light and further study to control this affect needs to be done as a future work. This indicated that scattering was the dominant mechanism for extinction. Understanding the affects of multiple scattering of laser light inside the ES PCL is essential to understand the complete analysis of PCL fiber with femtosecond laser. While some melting of fibers was observed on the edges of grooves ablated by the femtosecond lasers. the melting was minimal compared to that observed for grooves ablated with a Q-switched laser.4.4 Conclusions and future works The femtosecond laser was effectively used for ablation of meshes of electrospun PCL and PET polymers.1. 113 . The large widths of ablated grooves were attributed to material removal by optical radiation from laser-induced plasma. The extinction coefficient of PCL nanofiber meshes was measured and found to be high relative to the absorption coefficient of the bulk material. It was observed that the maximum width of the grooves was generally much larger than the laser focus spot size. a fact that was used to explain the observation that the estimated ablation thresholds of the fiber materials were somewhat larger than the bulk materials. Laser ablation resulted in uniform ablation grooves with width and depth controlled by femtosecond laser pulse fluence at constant scanning speed.
C. Electrospinning of silk fibroin nanofibers and its effect on the adhesion and spreading of normal human keratinocytes and fibroblasts in vitro. Characterization of nano-structured poly( e-caprolactone) nonwoven mats via electrospinning. Shin. Laurencin. Inai. S. Kwon. and I.5 References 1. Electrostatic spining of acrylic microfibres. Nanotechnology. 2003. Kotaki. 613-621. D. Lu. and F. I. R. P. H. 71-79. Aligned biodegradable nanotibrous structure: a potential scaffold for blood vessel engineering. 44: p. Biomaterials. D. Nanometre diameter fibres of polymer. Lee. Kim. Vacanti. B. 2004. Terai. Colloid Interface Sci. 10. 4. C. Lannutti. 114 . M. Electrospun nanofibrous structure: A novel scaffold for tissue engineering. 2002.E. 25(1289-1297). 26(37-46). 9. Res. W. Material science and engineering C. H. Li.Reneker. A biodegradable nanofiber scaffold by electrospinning and its potential for bone tissue engineering. 2007. T. Polymer. 877-886. Nam. 7: p. and D. Aguilar. Xu. K. 3. Biomaterials..4. G. 1996. S. 6. 25: p. Y. Mesoscopic spatial designs of nano. 27: p. and W. Mao. D. Lee. 2005. Min. Ma. Baumgarten. Lee. and T. 504-509. Kim. Biomaterials. and S. 8. 2005. 1971. H. Yoshimoto.and microfiber meshes for tissue-engineering matrix and scaffold based on newly devised multilayering and mixing electrospinning techniques. 7. 7642-7649. S. 2004. Electrospinning for tissue engineering scaffolds. Farson. 2. Chun. Y. Ramakrishna. Tomasko. J. Chen. and D. C. 36: p. Matsuda. T.. 24: p. M. R. Biomedical Mat. Reneker. Tuan. Khil. Direct micro-patterning of biodegradable polymers using ultraviolet and femtosecond lasers. Ra. Y. produced by electrospinning. Kidoaki.1. 216-223. Caterson. Lee. 2077-2082. 60: p. 5. Biomaterials. and S. Biomaterials. J. Y. 26: p. 2003. 1287-1294. Ko. and J. Park.
19. K. Surface Sci. Am. 26: p. A. Teoh. Plasma process polym. 21. Quant. P. and M. 168: p. Kuper. Electronics. and S. Advances in polymer science. 525-546. 14. and P. 555-560. Kruger. Bonse. J. and M. The role of plasma in ablation of materials by ultrashort laser pulses. B. Surface Sci. 154: p. D. Choi. Stuke. Pogue. Ashkenasi. Femtosecond pulsed laser ablation and deposition of thin films of polytetrafluoroethylene. Bovatsek. 17. 2003.. Kilmentov. M. J. Prokhorov. Sci. Soc. Salle. Pivovarov. B. Appl. Lan. Semerok. 115 . Kononenko. Scattering. S381-S383. Vendan. 247-289. 2005. 1987. T. P. 31: p. 763-769. A .Mater. and D. Processing. G. 44: p. Femtosecond and picosecond laser microablation: ablation efficiency and laser microplasma expansion. Appl. B: Photophysics Laser Chemistry. Phys. and D. 4-6. Kuper. 2713-2727.11. S. Phys. 2001. D. Femtosecond UV excimer laser ablation. Dehghani. Ultrashort pulse laser interaction with dielectrics and polymers. 48: p. Molian. Kautek. K. Ablation of Polytetrafluoroethylene (Teflon) With Femtosecond UV Excimer Laser Pulses. P. 2005. B. Lippert. Kruger. Womack. 12.. Biomaterials. O. Breitling. Appl. H. Farson. M. J. T. Kautek. 22. H. Meynadier. Lett. Arai. Waterman. M.. Gobert. 54: p. M.. Goh. 2: p. Biol. A. 16. 1999. 2430-2441. 199-204. 69: p. B. Hong. S. Stuke. 1989. Phys. 13. and W. Garnov. 2004(221). Paulse. Appl. S. J. 2000. 2007. J. Laser surface modification of poly(epsilon-caprolactone) (PCL) membrane for tissue engineering applications. A. and F. S. 20. 22(11): p. S. S. and extinction by thin fibers. Med. Brooksby. Konov. and W. The effects of internal refractive index variation in near-infrared optical tomography: a finite element modeling approach. Phys. and A. 15. Baudach. Ultrashort pulse laser ablation of polycarbonate and polymethylmethacrylate. 2005. Applied Optics. Z. V. Direct-write patterning of ITO film by high pulse repetition rate femtosecond laser ablation. Perdrix. 378-382. absorption. 2004. 18. Petite. Wang. Tiaw. Interaction of photons with polymers: From surface modification to ablation. Dausinger. Appl. Opt.. Vishwanath.
5]. Laser ablation with a scanned focus spot (direct-write ablation) offers important advantages over these other material removal processes including flexibility and speed [6. Microfluidic architectures require precise fabrication control and chemical wet etching. 7]. Due to this and other ablation characteristics. reactive ion etching (RIE) and plasma processes have been used as precision glass machining processes in clean room environments [4. 7-9].1 Introduction Glass dielectric materials have been widely used in many biomedical device applications due to their desirable thermal properties and chemical stability [1-3]. As a non-clean room process. Due to limitations of regenerative amplifiers which require the amplification time. it also provides a flexible way to fabricate 3-dimensional patterns by varying the beam scanning speed during ablation while clean room process requires highly controlled environment and maintenance cost. It also does not require a clean room environment. 3. Part of the reason for this low heat-input machining is the short time for conduction of energy from the laser interaction site by thermal diffusion. femtosecond laser machining has been found to produce superior results to conventional laser micromachining [1. focused ion beam (FIB). 116 .4.2 Laser machining of dielectric materials for biomedical applications 4. Laser ablation by pulses with duration on the sub-picosecond and femtosecond time scales can remove materials with lower residual thermal effects than nanosecondlong Q-switched laser pulses.2.
high energy femtosecond lasers provides the ability to perform ablation with pulse rates in the MHz range. m and e are the mass and charge of an electron . The peak power in femtosecond laser ablation reaches intensities where multiphoton ionization (MPI) and/or tunnelling ionization (TI) can create high electron densities and localized plasma formation by avalanche multiplication leads to permanent damage and ablation results . In transparent dielectric materials. The mechanism of dielectric material breakdown starts with the excitation of valence-band electrons by the incident laser. γ = ω mcε 0 E g e I = ω F 2Eg (4. positive ions are ejected in a process which is known as impulsive coulomb explosion (ICE) [13-15]. the pulse repetition rate of femtosecond lasers was limited to several kHz. Then. I is the laser intensity. high energy intensity is required.1) where ω is the laser frequency.2. The Keldysh number can be used to differentiate whether the photoionization is by MPI or TI [16-19] and can be calculated as Eq. (6) . The mechanisms of laser-induced damage to dielectric material in the ultrashort pulse regime have been studied by many researchers [10-12]. recent development of a high repetition rate.electro-optics device control timing and etc. Eg is the band 117 .. n is the refractive index of material. c is the speed of light. For the initial ionization. However. The non-thermal component of the mechanism is associated with emission of electrons which leads to softening and instability of chemical bonds. linear absorption of incident laser light is small.
but the fused quartz has higher thermal stability and bandgap energy (Eg). The bandgap energies of soda-lime glass and fused quartz are 5eV and 8. the Keldysh number is plotted with respect to applied laser intensities. respectively [23-25].11. Multiphoton ionization occurs when Keldysh number is γ > 1.gap of the material. Figure 4. while tunnelling ionization occurs when γ < 1 [17. With laser intensity and material properties. 118 . ε0 is the permittivity of the free space. Shaded area represents the experiment region.9eV. we calculated the Keldysh number to study the initial breakdown mechanism of dielectrics. 4. In Fig. and F is the electric field strength. 21] Both soda-lime glass and fused silica are transparent in the visual spectrum .11 Keldysh parameters for soda lime glass and fused quartz.
The laser intensity used for ablation is between 1.06 x 1014 W/cm2 and 3.5 x 1013 W/cm2 so the Keldysh number was below 1 for both soda-lime and fused quartz materials at the upper end of the ablation range but above 1 at the lower end of the machining range. which provides an opportunity for an interesting comparison on the materials affects for these two cases. at low pulse overlap. An “incubation” term is used to explain the effectiveness of lowering the ablation or damage threshold [26. the absorptivity will increase as surface morphology and color center formation of the material changes. Thus. the ablation threshold of ITO approached the single pulse ablation threshold fluence of 0. the ablation thresholds approach the ITO damage threshold fluence of ~0. One drawback of ultrafast pulsed laser ablation is a tendency to produce rougher machined surfaces compared with non-pulsed processes  due to the removal of discrete volumes of material. In our previous work. we can conclude that the initial breakdown of dielectric material in this experiment was be due to tunnelling ionization for some experiments but multiphoton ionization for others.25 J/cm2 reported elsewhere . The decrease of the ablation threshold is attributed to increase absorption due to accumulation of damage or defects from individual pulses. we study the effect of laser focus variables on surface roughness channels on soda lime glass and fused quartz 119 . The surface roughness can be reduced if lower laser fluence is used but at the expense of reduced ablation rate. In this work. 27]. With the damage occurring at these nonlinear absorption mechanisms.7 J/cm2 whereas at high pulse overlap.
A Gaussian radial fluence profile F(r) for an axisymmetric focused laser spot is ⎛ − 2r 2 ⎜ F (r ) = F0 exp⎜ 2 ⎝ w0 ⎞ ⎟ ⎟ ⎠ (4.2. the ablation threshold fluence.and surface smoothening process by applying a poly-hydroxyethyl methacrylate (HEMA) coating for the reducing of the surface RMS roughness. the diameter D of the area near the centre of the beam from which material is removed becomes 120 .2 Analysis of Laser Ablation Results needed for analysis of laser ablation are summarized below for convenience [14. 28].2. F0 is related to pulse energy Ep as F0 = 2E p 2 πw0 (4. Single pulse ablation experiments using known pulse energies allow this ablation threshold fluence Fth to be determined.3) If F0 > Fth .2. It has been found that many materials have a distinct power density threshold for ablation by femtosecond laser pulses.2) where w0 is Gaussian beam radius and F0 is maximum fluence. 4.
pitch p is defined as the distance between the centers of adjacent pulses.2. The horizontal pitch (corresponding to Y-axis direction. 4.12) was adjusted by setting the scanning speed s according to the relation p= s fp (4.5) where fp is laser pulse repetition frequency.⎛F 2 ⎜ D 2 = 2 w0 ln⎜ 0 ⎝ Fth ⎞ ⎟ ⎟ ⎠ (4. If there exists a threshold fluence. Fig. Fd.2. The longitudinal pitch (corresponding to the X-axis or scan direction. The slope of this plot allows calculation of the effective radius of the laser focus spot.4) A semi-logarithmic plot of D2 versus F0 allows Fth to be calculated as the x-axis intercept of a linear curve fit to the data. Fig. then Fd can be found by a similar procedure. but not for removal of material and F0 > Fd . Pitch was varied for different experiments but longitudinal and horizontal pitches were always set equal. 121 . 4.12) was controlled by the motion control program. for visible damage. For scanned ablation.
2.6) results in s fp (4.6) LO = d − (4. The focus spot overlap for adjacent pulses when the spot is scanned at a constant speed is illustrated in Fig.12 Definition of focus spot overlap LO.2. pulse repletion rate.2. both damage and ablation threshold fluences are less than for N = 1 and there is progressively greater reduction in thresholds with increasing N. focus spot diameter. This effect. The scanning speed.12. 4. Combining Eq. and focus spot overlap are related each other.s (4.5) and (4. termed incubation .Figure 4.7) If N > 1 laser pulses are incident on a location. focus spot overlap LO can be expressed as LO = d − p where d is focus spot diameter. is due to accumulation 122 .2. From the geometry. The distance of intersection between the focus spots of adjacent laser pulses is referred to as the focus spot overlap.
8) where N is number of pulses incident on a given location and ξ is called the incubation factor.3 Experimental Apparatus and Materials Our experiments used commercial Ti:Al2O3 femtosecond laser systems (CPA 2110 and CPA 2161.13 contains a detailed diagram of the optical beam delivery system. 4.2. Fig. The emitted laser beams had a central wavelength of 775 nm.2. The affect of incubation on ablation threshold can be quantified by a relationship of the form Fth ( N ) = Fth (1) N ξ −1 (4. Clark-MXR) The two laser systems had identical characteristics except as noted. pulse duration of 150 fs and pulse repetition frequencies of 2 kHz (CPA 2110) and 3 kHz (CPA 2161). 123 .of damage from individual pulses. 4.
0 μJ to 3. too large for precision machining.75 μJ by thin-film polarizing beam splitters (PBS) and a λ/2 waveplate.Figure 4. To set the 124 . Pulse energy ranged from 2. The laser beam was linearly polarized with polarization axis generally maintained normal to the scan direction. BD: beam dump) The un-attenuated output of the lasers had pulse energy of over 0. The pulse energy was attenuated to the range from 50 nJ to 3. For these experiments.75 μJ for channel fabrication and 667nJ for surface roughness and coating experiments. SHT: shutter. the beam polarization was rotated with a λ/2 waveplat so its axis was normal to the travel direction.5mJ. Some experiments were done to study the effect of polarization on the laser ablation.13 Block diagram of beam delivery system setup (PBS: polarizing beam splitter.
40 and the NA of the 50X objective was 0. an M2 quality parameter of 1. Focus spot radius and can be calculated by the relation w0 = 1 π M2 λ N/A (4. The NA of the 20X objective was 0.42.3 and was focused with a 20x microscope objective (M Plan Apo NIR 20X. The beam had a diameter of 5 mm.2.pulse energy for the experiments. laser output power was measured at the output of the focus optic with a power meter and the power divided by the pulse repetition frequency. Mitutoyo) and a 50x microscope objective (M Plan Apo 50X NIR.10) . Thus.9) The diffraction limited focus spot diameter d0 (twice the radius) was calculated to be 1.2.5 µm for the 50X objective lens. pulse energy of 667 nJ corresponds to average power of 2mW at 3kHz pulse repetition frequency. The propagation of higher-order Gaussian beams characterized by a beam quality factor can be modelled by the relation ⎛ M 2λz ⎞ w( z ) = w0 1 + ⎜ ⎜ πw2 ⎟ ⎟ 0 ⎠ ⎝ 125 2 (4.6 μm for the 20x objective and 1. Mitutoyo) as noted.
This spot size was used for the subsequent channel machining experiments. the focus lens was moved away from the material surface by 12. The system was equipped with a coaxial vision system for adjusting the focus of the laser beam relative to the material surface. and λ is wavelength .5μm encoder resolution on each axis to move the sample relative to the focus point (X.3 for soda-lime glass and fused quartz. respectively. For the channel machining experiments. Soda-lime glass substrates (12-544-1. 126 .2 and 4.5μm from the focus location to yield a focus spot diameter of 10µm.Y axes) and to position the focus optic relative to the sample surface(Z axis). Laser ablation was done on glass and quartz substrates. These and other material properties are listed in Tables 4. The laser system used a precision linear motor-driven rectilinear motion system with 0. Fused-quartz is expensive relative to soda-lime glass but is advantageous for some applications because it has higher softening temperature (1683°C vs. z is propagation distance from focus location.6mm (1/16”). The soda-lime glass was selected because it is widely used for biological applications at affordable price. Technical Glass Products) had a nominal thickness of 1. 494°C).where M2 is the beam quality factor. Fisher Scientific) had a nominal thickness of 500μm and fused quartz substrates (GE 124.
2 SiO2-72%. CaO-6.48 g/cm3 91./ 12-544-1 500μm 720C 2.50% 1.07nm 0.3%. NaO-14.3% Table 4.Supplier/Model Thickness Softening Temp Density Light transmission (VIS) Refractive index Poisson's ratio Chemical composition Fisher Scientific Inc.4% MgO4.517 @ 546.2 Material properties of soda-lime glass  127 .
4.6.2. F<0.1. the slope equals 2w02 and the effective focus spot diameter is 4µm./ GE 124 1.2g/cm3 93.6 Table 4.4. Li<0.2.4).Supplier/Model Thickness Softening Temp Density Light transmission (VIS) Refractive index Poisson's ratio Impurities (ppm) Technical Glass Products Inc. (4.01. the selected speed was high enough to separate the ablation cavities produced by individual pulses. Ca<0.0% 1. This diameter was used to convert pulse energy to fluence and the X intercept of the squared ablation diameter versus fluence plot in Fig. 4.4585 0.14(a). From Eq.3 Material properties of fused-quartz  4.17 Al<14. Ti<1.8.6mm (1/16”) 1683C 2.14(b) was used to calculate ablation 128 . Zr<0. the slope of a line fitted to the data was 8. The average diameters of the ablation cavities were measured from microscope images and the effective focus spot diameter and ablation threshold and were calculated. From the plot of squared ablation diameter versus energy shown in the Fig.4 Ablation Threshold Measurement Single pulse ablation features were generated with the 20X focusing objective lens while scanning at 30 mm/s.
(b) fluence.14 Single pulse ablation diameter as a function of (a) pulse energy.00 12. 34].00 0.00 6. 1.00 0.00 12.0J/cm2 depending on the types of glass [33.076 and the corresponding ablation threshold is 2.10 Energy (μJ) 1.00 (a) 14.00 4.079 10.0 100.0 10.00 8.00 2.0093Ln(x) + 15.00 0.7.00 6.00 8. 129 .42 J/cm2.00 4.00 2.7J/cm2 – 4.threshold. This measured value is in range with that found in other literature. The X intercept is -7.00 Ablation width2 (mm 2) y = 8.076 1.0 2 Fluence(J/cm ) (b) Figure 4.00 y = 8. 14.00 Ablation width2 (μ m 2) 10.0093Ln(x) .
possibly due to thermal radiation from plasma visible at the ablation spot. an experimental matrix was designed. Based on feasible parameter ranges determined from these preliminary experiments.Procedures and Results Initial ablation experiments used the 2kHz pulse repetition rate and 20X focusing objective and spanned a wide range of laser fluences and scanning speeds.2.5 Microchannel Ablation . argon gas was flowed over the ablation site at a rate of 39 x 10-6 m3/s to convect ablated vapor and debris away from the channel before condensation or redeposition could occur.4. Cracking and irregular glass surfaces were observed at the higher fluences. To decrease redeposition. 130 . It was also noted during the initial experiments that some ablated material re-deposited in channels.
0 2.75 3.0 7.0 μJ and 3.00 3.00 Pitch (um) 2.50 2.50 2. Laser pulse energy (Ep) was varied between 2.75μJ and pulse pitch (p) from 2.50 2.00 2.5 10.0 7. A channel width of 200 μm was targeted for these experiments and the number of adjacent scan passes was adjusted to achieve the 131 .75 3.5 5.00 3.5 5. These pitches corresponds to scan speed (s) from 5 mm/s to 20 mm/s and the pitch in the lateral direction (normal to scan direction) was set equal to the scan direction pitch.5 5.4 Experimental parameter settings As shown in table 4.75 3.5 10.0 2.Sample ID #1 #2 #3 #4 #5 #6 #7 #8 #9 #10 #11 #12 #13 #14 #15 #16 Ep (uJ) 3.0 2.5 μm to 10 μm.00 2.0 7.0 Table 4.00 2.00 2.0 7.5 5. ablation experiments were performed at 16 different parameter combinations.75 3.5 10.50 2.4.00 3.5 10.
the depth of ablation also varied for different experimental settings. 4. To achieve reasonable channel depths. Focus location was adjusted downward by a 5 μm increment after each layer to maintain approximately constant focus spot size.15 Ablation pattern with respect to laser fluence and scanning speed 132 . each experimental parameter combination was repeated to ablate 5 layers of material in the depth direction.target channel width in accordance with the pitch for that experiment. Figure 4. Since ablation depth varied with pulse energy. Sample results from the ablation experiment are shown in the macroscale scanning electron microscope (SEM) images in Fig. which compare the overall affects of pitch and laser pulse energy.15.
4.As expected. Scanned profiles of channels ablated at the low and high energy conditions are shown in Fig.16 Ablation depth as function of pulse energy for channels ablated at scan speed of 5 mm/s (2. As shown in the figures.16. 30 25 Depth (um) 5 Passes with 2. Veeco) to quantify ablation depth and roughness. the channels at the smallest pitch had the smoothest surface and the channels were scanned across the width by using a stylus profilometer (Dektak 3.5 µm pitch) 133 . larger laser pulse energy and smaller pitch (larger overlap) ablated the material to a larger depth.5um pitch 20 15 10 5 0 1 2 Energy (uJ) 3 4 Figure 4.
Sample #13 5 Depth (um) 0 -5 0 -10 -15 -20 -25 -30 X-direction (um) 100 200 300 400 (a) Sample #1 5 0 Depth (um) -5 0 -10 -15 -20 -25 -30 X-direction (um) 100 200 300 400 (b) 134 .17. A plot of microchannel depth as a function of pulse energy is shown in Fig.It is evident that the channel ablated at lower energy is smoother but shallower than the channel ablated at higher energy. 4. More ablation experiments were done at lower energies of 1.75µJ to further study the affect of ablation energy on depth and roughness.5µJ and 1.
5µJ but there is not any noticeable deviation from the linear trend in roughness at energies corresponding to 135 . This observation is notable since the Keldysh parameter crosses over the boundary from tunnelling ionization to multiphoton ionization at approximately 1. (b) 3.75µJ.18. 4. The roughness monotonically with energy over the range studied and the decrease was linear in the range from 2.5µm pitch (75% overlap).5µJ to 0.5µm pitch (25% overlap). 2. (c) 7. 2.5µm pitch (75% overlap).17 Profilometer scan across channels ablated at pulse energy and pitch (overlap) of (a) 2. The depth increased nearly linearly with energy over the range tested although.00µJ.Sample #11 0 -5 0 100 200 300 400 Depth ( μ m ) -10 -15 -20 -25 -30 X-direction (μm) (c) Figure 4. The peak to peak roughness Rt of the bottom of the channel was also measured from this scan data and is plotted in Fig.667µJ.
In previous work cited earlier.5 2.0 0.18 Peak-peak surface roughness versus pulse energy at a pitch of 2. It is noted that pitch and focus spot overlap are directly related to each other and also to the number of pulses incident on a point on the material surface within the scan pattern.0 2.5 1.0 Energy (μJ) 4. the decrease of ablation threshold with number of incident pulses measured for a linearly-scanned focus spot and was found to obey the same incubation relationship as for the fixed fluence.0 Figure 4. RM S roughness (μm ) 3.0 1.5 0. This is certainly a reasonable observation since one would expect the fact this ionization occurs at all to be more important than the mechanism by which it occurs.5 µm.the different ionization mechanism. From this one could conclude that the ionization mechanism has no affect on the ablation roughness for the material being used in these experiments.0 0.0 3. A situation that is similar to linear ablation occurs when a focus spot is scanned over an area 136 .0 1. fixed location experiment.0 2.
with a fixed pitch.2. If a point is said to receive a laser pulse if it lies within one Gaussian radius of the location of the beam center when the pulse occurs. xi and yi are the relative central position of the ith and jth pulses and w0 is the radius of focused spot.2) is Gaussian. 1. the accumulated fluence is a function of location and Gaussian distribution function.2. some locations within the area scan pattern are exposed to energy from more than one pulse.28 for 75% overlap. and 6. The importance of focus spot overlap and incubation was further analysed for 6 of the experimental conditions that showed significant and repeatable effects for pitch. 25%. The accumulated fluence as a fraction of the single pulse peak fluence F/F0.0 at 0% overlap. Since the intensity distribution as described in Eq.04 at 25%. The accumulated fluence increased from 1. the total accumulated fluence was as high as 15. 50% and 75% focus spot overlaps. 137 . In terms of the experiments. The total amount of energy received depends on the location and focus spot overlap. y ) = i = −∞ j = −∞ ∑ ∑F e 0 ∞ ∞ ( −2 (( x − xi ) 2 + ( y − y i ) 2 ) / w0 2 ) (4. (4.42J/cm2.11) where i and j are integer numbers. 4.2J/cm2. 1.19 for 0%. was calculated and results are shown in Fig.62 at 50% overlap. for F0 = 2. One can calculate the accumulated fluence at a given point by the scanned laser focus point as F ( x.
max accumulated fluence factor: 1.4 0. max accumulated fluence fraction is 1 -10 Y location (um) -5 0 5 10 -10 -5 0 5 X location (um) 10 Fluence enhancement 1.8 0.5 1 0.5 0 -20 -10 0 X location (um) 10 20 (b) 25% overlap.04 138 .-10 Fluence enhancement Y location (um) -5 0 5 10 -10 -5 0 5 10 1 0.6 0.2 0 -20 -10 0 10 20 X location (um) X location (um) (a) 0% overlap.
(d) 75%.5 1 0.5 0 -20 -10 0 10 X location (um) 20 X location (um) (c) 50% overlap. max accumulated fluence factor: 6. The right plot is the cross section of fluence enhancement variation along with Y location =0μm.2 Y location (um) Fluence enhancement -10 -5 0 5 10 -10 -5 0 5 10 1. 139 .28 Figure 4.62 Fluence enhancement -10 Y location (um) -5 0 5 10 -10 -5 0 5 10 6 4 2 0 -20 -10 0 X location (um) 10 20 X location (um) (d) 75% overlap. (c) 50%.19 Accumulated laser fluence (a) 0%. (b) 25%. max accumulated fluence factor: 1.
Figure 4. 75% overlap.17(c) is close to that shown for the 25% overlap case in Fig. 4. the period of corrugation visible in the left half of the channel profile in Fig. The match is not perfect because the phase of pulses if adjacent linear scans are subject to variation because of differences in starting time of the scan motion relative to the laser pulse waveform.19(b). Roughness produced by this mechanism would be expected to have a periodicity similar to that of the intensity enhancement plots in Fig. The period of the minima in both cases is about 7. 50%. focus spot overlap. Indeed.5µm.These plots give an indication of the roughness that may be expected simply due to intersection of ablation cavities produced by individual laser pulses. 140 . Arrows show 25%. 4. 4.20 Accumulated fluence as a fraction of single pulse peak fluence vs.19.
At the 75% overlap condition which produced the smoothest channels.2. (4.21. Although this value should be related to the magnitude of incubation at these points. 4.12) and results are plotted in Fig.29. Another advantage of higher overlap can improve the uniformity of laser fluence.12) 141 . 4. As shown in Fig. 4.The maximum accumulated fluence at points along a linear scan relative to the pulse fluence is plotted as a function of focus spot overlap in Fig.20. δ= ( Fmax − Fmin ) Fmax (4. there is a laser fluence variation at lower overlaps. it is still an open question as to the effect of this accumulated energy on ablation threshold and further study is needed. The variation can be as Eq. the maximum accumulated energy is over 6 times the pulse energy at some points along in the channel.2.
The P-polarization is defined as the linear optical polarization axis being aligned with the scanning axis. It was observed that a P-polarized light laser machining approach was a more effective method for ablation.21 Fluence variation vs. overlap As we can see in Fig. it is obvious that the fluence variation at 75% overlap is very negligible. So. 4. Separate experiments were conducted to quantify the effect of polarization angles on the quality of channel ablation. The polarization was changed using a λ/2 waveplate.2%) or higher. and S-polarization as the linear optical polarization axis being perpendicular to the scanning axis. but S polarization resulted 142 .Figure 4.21. the variation of fluence at different locations is almost negligible once the pulse overlap is kept 60% (less than 0.
channels with a more square profile as shown in Fig.22. 4. Similar results were also reported during a high aspect ratio trepanned drilling process . Presumably. Figure 4. 143 .22 Optical images of the channel quality and shape with P-polarized beam (left) and S-polarized beam (right) on fused quartz. improved machining occurs with P-polarization because absorption is more effective on the front wall of the channel formed as the beam scans over the material.
The laser pulse energy was 667nJ which corresponding to 2mW at 3kHz of pulse repetition rate and scanning speed and lateral offset were adjusted to produce 25%. material is removed as a series of partially intersecting cavities so roughness is minimized by combinations of high focus spot overlap and low laser fluence.4. both channel quality and productivity can be optimized.6 Surface roughness and HEMA coating process By the nature of scanned. 50% and 75% focus spot overlaps. In this section. The use of a coating also enhances design flexibility since the coating may be functionalized to change hydrophilicity or to reduce nonspecific binding of proteins. higher focus spot overlap and low laser fluence both decrease the material removal rate of the ablation process. By reducing the surface roughness of channels with a coating polymer. However. HEMA has been used to bond complex microfluidic devices and for surface modification  in devices produced by photolithography. Five micron areas were selected randomly for roughness 144 . making it a good selection for surface modifying devices for biological. diagnostic. we describe the surface roughness improvement achieved with HEMA coatings. using the 50X infinity corrected microscope focusing objective lens with a defocused 5μm focus spot diameter.2. HEMA has good optical properties as well. and other microscale applications that involve optical sensing or photochemistry. Square 5mm x 5mm areas were prepared by laser ablation with a pulse repetition rate of 3kHz. Some redeposition of ablated material was noted in these large areas due to incomplete removal by the argon gas flow. pulsed laser ablation.
measurement, with smaller (1-2μm) subsets selected within the 5μm area also measured. Localized roughness much larger or much smaller than the overall average could easily be found and we selected area which can well represent the entire area of ablation. To coat channels with HEMA, we followed a procedure similar to that described by Lai et al. . The procedure is described in Fig. 4.23. Hydroxyethyl methacrylate (HEMA, monomer) and Phosphine oxide, phenyl bis (2,4,6-trimethyl benzoyl) (Irgacure 819 photoinitiator) were purchased from Aldrich (Milwaukee, WI). After ablating a channel, a lid was bonded to the channel. Self-adhesive polymer films, thermally bonded polymer films, polymer films bonded with UV-cured adhesive or unbonded lids all may be used with this process. HEMA monomer mixed with initiator was then loaded into the channel by vacuum suction. Next, either positive or negative (vacuum) gas pressure was reapplied until the gas bubble penetrated the HEMA along the entire length of the channel. By adjusting gas pressure, coating thickness can be adjusted and the detail quantification is not made in this article. Gas pressure was continuously maintained as the HEMA was cured under high intensity UV radiation resulting in a uniform, smooth coating. The improvement of surface coating is observed by optical microscope shown in Fig. 4.24.
Figure 4.23 HEMA coating process
Figure 4.24 Optical images of HEMA coated channels (before and after)
To quantify the surface roughness improvement, we experimented HEMA coating on uncovered surfaces with 5mm x 5mm ablated area with different overlaps. To have uniform spreading of HEMA coating, one can apply spin coating or gas blowing method, and we used gentle gas blow after dropping HEMA on the surface. In this case, the thickness or uniformity control was much more challenging.
4.2.7 Surface roughness characterization
As shown in Fig.s 4.25 and 4.26, the surface roughness of soda-lime glass and fused quartz was measured with Atomic Force Microscope (AFM) by Asylum Inc. Both in AFM and by standard brightfield optical microscopy, individual ablation spots could be seen on the 25% overlap sample. Also in the case of 25% overlap, the depth of the holes could not be directly measured by AFM because the probe tip could not in all cases read the bottom of the hole. The pyramidal geometry observed at the bottom of the ablation holes is an artifact of the tip geometry and is not a read feature of the surface. The RMS value of surface roughness was calculated within the custom AFM interface provided by Asylum, Inc. developed based on the Igor Pro data analysis platform.
Figure 4.25 Surface roughness data for different overlaps. The numbers inside parenthesis is the scanning area. E=667nJ, material: soda-lime glass
Figure 4.26 Surface roughness data (E=667nJ and different overlap) for fused quartz
The 25% overlap data represents the average roughness of areas that include holes; if holes were excluded, much lower values could be measured. The 50% and 75% data contains measurements of 5μm areas as well as a selected of smaller areas to capture the roughness variations within certain localized features. We speculate that the lower roughness for 75% overlap data is due to increased ablation and isotropic re-deposition of
or may be stress in the HEMA due to the high intensity 149 . Two vertical scales are provided: The upper row shows near proportional vertical scale (5μm) hile the lower row shows a highly magnified vertical scale (500nm) to see features. AFM scans show blob-like features indicative of re-deposition on all images. 4. The HEMA coated samples show a uniform.27 Surface roughness data (E=667nJ and different overlap) for HEMA coated soda-lime glass. thin coating of HEMA over the surface of soda-lime glass shown in Fig. Figure 4.27. Large scale undulations may represent the contour of the underlying surface.some ablated material. with the 75% overlap images showing the most numerous.
Figure 4. Because the undulations are less noticeable in the 75% sample.28.cure. The surface roughness comparison is summarized in Fig. 4. The results displayed show suitable smoothness for most biological and diagnostic applications.28 Surface roughness data of soda-lime glass and fused quartz 150 . it is likely at least some of the undulations are due to the underlying surface.
A systematic approach was applied to find an optimum window of ablation process parameters. Laser pulse energy was varied between 2.75μJ and the roughness of surface was measured to be between 295nm and 731nm RMS depending on both pulse energy and overlap.0μJ and 3.8 Conclusions and future works In this paper. The characterization of coating thickness and surface roughness vs.2. vacuum pressure needs to be further investigated. 151 . Surface roughness was improved to around 100 nm when the pulse energy was decreased to 667nJ. Coating of carbons or hydrogel on quartz channel is planned to investigate its functionality. Other bio-compatible polymers or functional molecules coating could be used as polymer coating material and further study is needed.4. HEMA coating was effectively used to decrease the surface roughness to around 10 nm. we are presenting a non-clean process for dielectric material ablation by using femtosecond laser system. The developed technique was used to fabricate for microfluidic channels on dielectric materials (soda-lime glass and fused quartz) in conjunction with polymer coating for surface roughness modification.
3. M. D. 6. 506-513. M. Focused ion beam microfabrication in Foturan(TM) photosensitive glass.A. 2005. Gambino.. 33(10): p. Prewett. Reider. J. Zoppel. and M. Wei. Crack-free direct writing on glass using a low power UV laser in the manufacture of a microfluidic chip. 21: p. D. Microeng. G. and K. Wu. 27-40. 152 . Sampath.D. J. Ultrafast laser micromachining and patterning of thermal spray multilayers for thermopile fabrication. IEEE journal of quantum electronics. 2. 2003. K. P. Venkatakrishnan. J.. Longtin. Tankiewicz. 1706-1716. G. I. Docker. 5. L. Farsari. 44(5): p. I. 7. Microeng. 775-779. 2000.. Femtosecond laser ablation for microfluidics.. 2545-2549. Optical engineering. P. Efficient femtosecond laser micromachining of bulk 3C-SiC. and Y. Micromech. O. 4. Liu. S. C. Micromech. D. Tech (B). Chuang.4. Sci. Gomez. 21(6): p. 2007. 1997. 2005. Micromech. J.2. 15: p. 2002. Anderson. Q. Chiu. Electrophoresis. and G. Li. D. and T.S. Esashi. T. 12: p. Jiang. Micromech.9 References 1. 9. Laser ablation and micromachining with ultrashort laser pulses. J. Smooth surface glass etching by deep reactive ion etching with SF6 and Xe gases. Mourou. J. 15: p. 17: p. J. Micromech. Fabrication of microfluidic systems in PDMS. Cheng. Ozaita. and L. J. H. 14: p. and G. 115-119. and M. X. Lizuain.F. C. Young. P.G. J. S. Microeng.. Whitesides. McDonald. 1147-1156. S.. 2004.Chen. Femtosecond laser ablation of thin films for the fabrication of binary photomasks. Microeng. Lim. 8. Schueller. Du. Microeng. J. Anthony. 051105:1-8.Y. 2005. and Fotakis. and R. Vac. 1786-1789.
C. Stuart.M. Prade. Tunneling ionization in the multiphoton regime.. 41(9): p. L. Schaffer.M. X. Technol. 1991.. E. and S. 53(4): p. Multiphoton ionization of atoms. Phys. Farson. Ablation of solids by femtosecond lasers: Ablation mechnasim and ablation thresholds for metals and dielectrics. 64(23): p. a. Meas. Sudrie.. 17. Brodeur and E. C. G. D.. and A. Choi. 2002. 89(18): p. 1784 . 2002.5052. M. P.R. B. Franco. 9(3): p. 16. Applied surface science.. Journal of Laser Applications.1372. and V. 1994. 1996. Luther-Davies. T. 275-287. D. 2007.. Gamaly. Lamouroux. J. .. Mysyrowicz. J. and G. and T. J. Physical Review A. in Physics. Gibson.C. 19. B. 54(10): p. B. Mainfray. H. 14. 186601.1794. MA. 15. Lannutti. 2004. Du. Physical Review B. Coyne. 223: p. O'Connor. Structuring electrospun polycaprolactone nanofiber tissue scaffolds by femtosecond laser ablation. A. E. G. S. Tzortzakis. J.J. B. 1990. Herman. 13. M. 153 11. p. Mannion. Laser induced breakdown and damage in bulk transparent materials induced by tightly focused femtosecond laser pulses. The effect of damage accumulation behabiour on ablation thresholds and damage morphology in ultrafast laser micro-machining of common metals in air. and C. Appl. Interaction of femtosecond laser pulses with transparent materials. A. Nanosecond to femtosecond laser-induced breakdown in dielectrics. 1749-1761. Nam. Femtosecond laser-induced damage and filamentary propagation in fused silica. G. 3071-3073.. Squier. Glynn. G. Physical Review Letters. 18. Reports on progress in physics. Mazur. Harvard University: Cambridge. 1333 . Tikhonchuk. A.L. 12. Mourou. 2001. 12: p. J. Feit. Rhodes..C. 2001. Physics of plasma. Lett. 173. Schaffer. 5049 . 949-957.L. Laser induced breakdown by impact ionization in SiO2 with pulse widths from 7ns to 150fs.10. Korn. Sci.
Ashkenasi. S397-S407. S. and F. B: At. and D. 1996. M.20. D. 96(98): p. G. C. Appl. Appl. M. and K. and S. Stoian. 150: p.K. Ashkenasi. Ashkenasi. Rosenfeld.. Richardson. D. 1999.. R. Yuan. Rosenfeld. 20: p. 28. Ilkov. Phys. Corkum. 2007(101): p. 39: p. 23. Kruger. S373-S376. Mourou. 1307-1314. Cheng.. Choi. V. 2006. J. Physical Review Letters. 430-438.W. Surface Sci.. Applied Optics. Arai. Stoian. Ultrashort laser pulse damage threshold of transparent materials and the role of incubation.. 154 . Lorenz.): p. F. and A. 4076-4079. Opt. Journal of Applied Physics. 80(18): p. Ionization of atoms in the tunneling regime with experimental . J. M. 25. 1999. Direct-write patterning of ITO films by high pulse repetition frequency femtosecond laser ablation.K.. Phys. Ionization in the eld of a strong electromagnetic wave. J.F. Rajeev. Spielmann.T. J. 4005 . 24. Phys. L. A.S. M. 183: p.evidence using Hg atoms. R.D. Krausz. D.a. 2007. Soviet Physics JETP. L. Keldysh. 1992. 21. 1965. 024901-1:6. Femtosecond laser deep hole drilling of silicate glasses in air.Atomic molecular and optical physics.. Strong field ionization inside transparent solids. Surface damage threshold and structuring of dielectrics using femtosecond laser pulses: the role of incubation. Kautek.4020. 22. 2001. Richardson. Femtosecond-pulse visible laser processing of transparent materials. Bhardwaj. and D. J.C. Lenzner. Journal of Physics B . 151-164. Shah. and D. Femtosecond Optical Breakdown in Dielectrics. Rayner. Bovatsek. J. 69(Suppl. Sartania. 25(19): p. Applied surface science. Mol. Z. Experimental and theoretical anlysis of direct-write laser micromachining of polymethyl methacrylate by CO2 laser ablation. W. 27. P.. 26. 29. P. 1998. A. A. H. a. Applied surface science.L. 101-106. Das.
563-567. Ostendort.. Lee. Ashmore. Stone.30. Microslide specification. Phys. J. A. A Packaging techinque for polymer microfluidic platforms.5323. A. 2007. Phys. and H. 96: p. 2007. D. 1447 1459. Jackson. 31. Ben-Yakar. Ben-Yakar. Thermofisher scientific. Technical Glass Products. Harkin. Lai. 33. J. C. G. 1999. Nolte. Byer. 132-142. 35. and H. 1175-1183. 2004. Chem. and R. New York: Adam Hilger. 34. 36. ed. Thermal and fluid processes of a thin melt zone during femtosecond laser ablation of glass : the formation of rims by single laser pulses. 2007. Alvenslebe. Polarization effects in ultrashort-pulse laser drilling.J. 1989. Sasnett. 40: p. Byer.. Propagation of multimode laser beam-the M2 factor. Fallnich. 68(6): p. Femtosecond laser ablation properties of borosilicate glass. Cao. D. Kamlage. Properties of fused-quartz. 2004. J. M. 76: p. Applied physics A . Appl. F. and L. Momma. A.materials science & processing. 155 . and P. S. S. Hall. C. 32. A. Anal. R. 5316 . X. Welling. Appl.
5]. Being a direct-write process.4. non-linear and multiphoton affects associated with the high energy density can cause absorption within a small volume of material at the focus . When an ultrafast laser pulse is focused inside of a transparent material. At lower energy densities. By translating the beam so that multiple pores are connected.3. and to store data inside of material . Precision bulk or internal machining of transparent materials is another interesting capability that is offered by these short-pulse lasers. microfluid passages can be machined inside of the material. Using 156 . Creation of internal channels for gas and fluid flow is important for fabrication of bio-medical devices. this technique of fabricating internal channels is useful for device prototyping and low-volume manufacturing. Channel fabrication with generally-available technologies requires multiple process steps: fabrication of external channels on a substrate as well as a cover plate and subsequent bonding of the cover to the substrate to form closed channels. absorbed laser energy may modify the optical or other properties of the material and such bulk material modification has been used to write waveguides [2.3 Femtosecond Laser Bulk Micromachining of Microfluid Channels in PMMA 4.1 Introduction Focused amplified femtosecond pulsed laser beam have very high energy densities that are widely-used for surface microstructuring of many materials. the material becomes vaporized and ionized by the laser energy and its expansion causes a small pore to form at the interaction site [1. If the absorbed energy density is high enough. 3].
In this paper. it is possible to directly fabricate internal channels.high energy femtosecond lasers pulses. Remaining particulate or condensed material can clog the channels. this process is limited to materials whose chemical etching rate can be increased by absorbed femtosecond laser pulses . However. A gas flow system was configured to partially-evacuate gas. although not nearly as extensively . At fluences below the threshold for plasma formation. 157 . internal channel fabrication by femtosecond laser irradiation followed by chemical etching has been recently developed and fabrication of very clean internal channels was demonstrated. The use of flowing gas for material removal is more convenient than liquid for practical reasons because it eliminates the need for leak-tight seals. preventing the creation of an open passage. In previous reports of internal machining in glass. eliminating the fabrication steps needed to cover an external channel. 7-9]. removal of debris from the internal channel as it is being machined is a challenge. vapor and fine debris from the channel as it was being created. an alternative method for the creation of microfluid channels inside of poly(methyl methacrylate) (PMMA) is described. At higher fluences. liquid-assisted material removal is complicated since it requires liquid-tight seals and fabricated channel sizes have been found to be limited . No matter what the material. internal micromachining of glass by femtosecond laser pulses has been investigated by several workers [4. However. creation of microfluidic channels in polymer by laser bulk machining has also been studied. liquid has been used to to assist in debris removal so as to prevent channel clogging. Because polymers are more useful for many biomedical microfluidic device applications .
the laser pulse interaction in a transparent material is non-linear and self focusing can be expected to decrease focus size below the un-affected dimensions . The generated plasma evaporates the material and high pressurized gas phase material expands and deforms the focused region creating a void.The femtosecond laser has unique advantages for material processing over nanosecond or longer pulsed lasers due to the material interactions associated with the high intensity that can be achieved when the beam is focused [12. Multiphoton processes decrease the effective size of the laser interaction zone even further since photon density is a very sensitive function of distance from the center of the focus. the intense laser energy is absorbed and forms a plasma inside the PMMA. resulting in void formation. Unfortunately. it appears that laser ablation by-products in the internal 158 . which will eventually make internal channel. At high intensity. when the pulsed laser energy is properly focused inside material. the irradiance incident on the surface of material is not high enough to damage the material while the high irradiance at the focus initiates absorption by generating free electrons via multiphoton ionization and tunneling ionization. By scanning the laser focus.and nanofluid channel fabrication in PMMA is desired in chemical and biomedical applications. the continued action of this pulsed ablation process can produce connected voids. Also. by these affects. the absorption volume is restricted to a small volume and the energy density in the absorbing material becomes so high that the temperature is above the ionization temperature . Thus. In our work. The induced plasma locally compresses the solid material. a capability for micro. 13]. When a femtosecond laser beam is focused inside PMMA material.
By analysis of decomposition by-products that evolve from the material during the incubation phase. 159 . some amount of volatile methyl-methacryalate (MMA) vapor is produced at both IR and UV wavelengths. cracks and rough internal surfaces can be observed in the internal channels. The gas phase material can be absorbed by surrounded material or can remain inside the channel. To fabrication high quality channels of maximum length. It is also useful to note that the ultraviolet (UV) laser ablation of PMMA has been extensively researched  and some of the results are pertinent to this work. Although composition of the evolved species depends upon wavelength. the reaction steps can be inferred. A notable feature of ablation at wavelengths longer than 198 nm is incubation – visible damage does not occur until a number of laser pulses have been applied at the same location. material removal is one essential task to be solved.channel lead to excess plasma pressure and as a result.
4.4. In the configuration shown in Fig. Figure 4.3. a flow channel (2 ~ 5 millimeters in diameter) is first formed.29.29 Concept of gas assisted material removal The laser ablated internal microscale passage is then created starting at the boundary of this flow channel. we note that the flow of a compressible gas can be modeled as incompressible flow if the pressure changes are less than 20% of mean pressure . To analyze this process.2 Concept of gas assisted material removal: The concept of gas assisted material removal has been developed based on the reduced static pressure generated by high velocity fluid flow. This assumption is valid in this application since the flow 160 .
but also monomer or oligomer forms of MMA. based on the considerable laminar flow velocity that is developed. Based on incompressible fluid assumption. However. although any remaining over-pressure due to the material vaporization will assist in its removal. The flowing gas can more effectively evacuate removed material particles once they reach the flow gas channel. It is reasonable to assume that the gas or vapor produced by the ablation quickly reaches ambient temperature induced air expansion. 161 . oligomer or monomer phase particles may not be as effectively removed by a vacuum pump system. A vacuum to assist in material removal could also be created at the channel entrance using a vacuum pump. The actual pressure development at the laser interaction site during the internal machining process is much more complicated than is suggested by the simple analysis due to generation and expansion of high temperature plasma. The gas phase material then can be effectively removed by pressure differential between a nearby gas flow channel and material ablation spot. It is believed that either the static pressure from a flow gas or low pressure created by a vacuum system could be somewhat effective for material removal since particles or gaseous products produced by ablation would be induced to flow toward the lower pressure region in either case. Femtosecond laser ablation forms not only gaseous products.gas is delivered at moderate pressure (345 kPa) and pressure in the fluid channel is on the order of but somewhat less than atmospheric pressure (101 kPa). Bernoulli's equation can be applied to correlate the pressure and fluid velocity in the channel.
The maximum output power of the laser was 1. The beam was focused for processing using a 10x microscope objective with numerical aperture of 0.6 W. Using two polarizing beam splitters and a rotatable half-wave plate to control the polarization angle of the linearly polarized beam. gas flow rate and focal spot scanning speed. the pulse energy was varied from 0.4.3 Experimental PMMA (Plaskolite™) plates having thickness of 2 mm were used for the experiments.8 mJ. The delivery pressure was maintained at 374 kPa and the flow rate was varied between 0. Compressed Argon was selected as a flow gas because of its inertness and ready availability.5. PMMA has mechanical properties and biocompatibility that make it useful for a number of biomedical device applications and its transparency allows it to be processed by internal machining. The beam quality factor (M2) of the laser was measured to be 1. The surfaces of the prepared hot embossed samples were also scratch-free.5 μm.3.6. Polished steel pins were used in a hot-embossing process to make smooth-walled gas flow channels. corresponding to an average pulse energy of 0.25 μJ to 1 μJ and scanning speed was varied 162 . Thus. the wavelength was 775 nm and the nominal beam size was 5 mm. the calculated minimum focus spot size is approximately 2.63 ml/s and 2. so that the laser beam could be focused inside material without interference from the surface. Experiments were carried out by varying laser power. To create channels.6 ml/s by a manual flow control valve. The laser used in the work was a Ti:Al2O3 chirp-pulse amplified laser which produced 150 fs pulses at a repetition rate of 2 kHz. the pulse energy could be attenuated from this energy to a minimum of 5 nJ per pulse.
internal channels were machined without venting. In subsequent experiments. which were done for comparison. In the first experiments. three different conditions were tested. Using the calculated focal spot size. 163 . the pulse overlap is calculated to range from 80% to 96%. pulse repetition rate and scanning speed. The second experiment was done with vent hole but no gas flow.2 mm/s to 1 mm/s.from 0. Figure 4. 4. pressurized gas was applied through the gas flow channel at various flow rates.30 Experimental setup To determine the effectiveness of gas flow for improving channel quality. A photograph of the experimental set up is shown in Fig.30.
4 Results and discussion Channels were first created using a single-pass scanning procedure. Single pass experiments: It was found that for most effective debris removal. 4. single-pass channels had to be created by initiating machining at a free surface and then translating the laser beam focus inward. Results from both procedures are described below. a multiple-pass procedure was also developed. the channel was easily closed after the focus had passed by. To make sure pores were connected each other to form a channel.3.31a) and machined starting from the wall of a vent hole that had no Argon gas flow (Fig. the pulse energy used to form these channels was relatively high and the scanning speed was low. the pore size was correspondingly decreased. Fig. the focal spot scanning speed needed to be adjusted depending on pulse energy.31b). It was observed that as laser pulse energy decreased. To prevent clogging of the channels by debris. 4. At higher pulses energies.4. Even with this procedure. when doing internal machining at low pulse energies. a continuous channel could be formed. 164 . Because the application required wider channels than could be created with this technique.31 shows representative internal channels fabricated without venting (Fig. 4.
In the second experiment.31 Channel fabrication without gas assistance (a): no vent. 4. It is supposed that this effect might be due to violent explosion of combustible methyl methacryalate MMA vapor that results from decomposition of PMMA.31 appear to be quite rough. Both channels have very rough walls and the former has extensive adjacent cracking. However. there were still a few clearly visible cracks adjacent to the channel.(a) (b) Figure 4. 165 .2 mm/s The surfaces of the channels shown in Fig. which appeared to be continuous but had very rough walls in the back-lit optical photographs. (b): vent but no gas flow. the vent hole was helpful to minimize the cracks. Laser pulse energy was 1μJ and scanning speed was 0.
With the Argon flow rate near optimum of 1. 4. but the results depended on flow rate.32a was made at a flow rate of 0. the channels were more often clogged by debris and their walls became rougher. 4. 4. increasing the gas flow rate from 0. 166 .6 ml/s degraded channel quality rather than improving it.31.3 ml/s. debris and gas/vapour removal was assisted by argon with a delivery pressurize of 345 kPa applied through the gas flow channel.65 ml/s (5 ft3/hr) while that in Fig. This feasible machining length is long enough for some microfluidic channel applications.6 ml/s (20 ft3/hr).32b was made at a flow rate of 2. By comparison of the results with those in Fig. Interestingly.In subsequent experiments. The channel shown in Fig. although longer channels may be desired for many purposes. At longer distances. it can be concluded that the channel shape was noticeably improved by the gas flow.65 ml/s to 2. the channels appeared to be clear of debris and relatively uniform for distances of approximately 5 mm to 10 mm from the vent hole.
Channels could be created using pulse energies of 250 nJ to 1. Multiple photons from femtosecond laser irradiation are strong enough to break carbon-carbon bonds and reactive gas such as oxygen could react with and facilitate removal of ablated materials. This situation is analogous to the RIE (Reactive Ion Etching) process which involves plasma generation with a reactive gas and surface material interaction .6 ml/s (20 ft3/hr) – Laser pulse energy = 1 μJ and scanning speed = 0.32 Channel fabrication with gas assistance at different flow rates. the etching might be more effective.65 ml/s (5 ft3/hr) right: 2. since the primary purpose of using the flowing gas is to create low pressure to assist in removal of photo-decomposed material 167 .2 mm/s. Left: 0. If reactive gas could somehow be introduced to the ablation site. Unfortunately.Figure 4.0 μJ and sc It was considered that that the channel length might be increased by using a reactive flow gas instead of argon.
4. Additional experiments were conducted to modify the process to produce larger channels. which is the same as could be achieved with the single-pass machining technique. A typical maximum feasible channel length with this method was found to be between 5 mm and 10 mm. Instead of using a one-dimensional linear scan. A photograph of the channels is shown in Fig. Multi-pass experiments: The diameter of the single-pass channels was smaller than desired for some applications.33. flow gas concentration at ablation site would be expected to be low and flow gas reactivity may not be a significant factor.from the ablation site. Both pitch and width of zigzag pattern were set to approximately 15 μm. 168 . Processing parameters such as laser pulse energy and scanning speed were fixed while only gas flow rate was changed to determine its influence on channel quality. a two-dimensional zigzag pattern was generated.
Figure 4. the samples were cut and channel cross sections were viewed with a SEM.3 ml/s (10 ft3/hr). 0. Although a channel cross section is not displayed.2 mm/s For a better measurement of channel profiles.6 ml/s (20 ft3/hr). The channel width varied between 15 μm and 25 μm. 1. 169 . 2. 4.65 ml/s (5 ft3/hr). As shown in Fig.65 ml/s gas flow resulted in best channel quality. The samples were cut approximately 5 mm from the inlet of the entrance where the channel was believed to be fully developed. optical photographs showed that the channel was not clearly developed at 2.6 ml/s of flow rate.34. Laser pulse energy = 1μJ and scanning speed = 0.33 Plan view of channels created with multiple passes– from the left 0.
Figure 4. right-1.65 ml/s.34 Cross sectional view of channels (left – 0. By comparing the Reynolds number and channel profile. 4. we can conclude that laminar flow (Re < 2300) region provided the best effects.35. To correlate the channel profile with flow rate. the Reynolds numbers were calculated and plotted in Fig.3 ml/s). presumably because of more effective ventilation. 170 .
8000 6000 4000 2000 0 0 1 2 3 Flow rate (ml/s)
Figure 4.35 Reynolds numbers in the flow channel for 0.65 ml/s, 1.3 ml/s, 2.6 ml/s
To better characterize the shape and roughness of the machined channels, some of the multipass channels were photographed after a fluorescent dye (Flouresin yellow) was flowed through them. Fig. 4.36 shows fluorescence background of the channels under UV light (365nm) using inverted fluorescence microscope (TE-2000S, Nikon).
Figure 4.36 Micrograph of internal channels before being filled with a liquid containing fluorescent dye.
The two bright lines indicate that the femtosecond laser machining created open passages, allowing the dye to flow in and change the fluorescence background of the ablated area. Fig. 4.37 show the fluorescence intensity difference before and after the channels were filled with fluorescence dye. It can be seen that the intensity increased, also indicating the channels are open. Although the distribution of dye is relatively uniform, indicating that the channel walls are probably relatively smooth, several pockets of brightly fluorescing trapped dye may reveal surface discontinuities.
Before FITC (acid-yellow) is added
After fluorescence dye is added:
Figure 4.37 Flourescense spectra confirming presence of yellow dye in a channel.
4.3.5 Conclusions and future work
A concept for evacuating ablation by-products during the fabrication of internal channels has been described and demonstrated. The static pressure of a flowing gas in a vent channel provided pressure difference for evacuation of small debris and vapour the channel. Experiments showed that the evacuation of gas from the channel was effective for improving channel quality. In the experiments, argon gas was used and the venting technique resulted in visually smooth channels with no obvious cracking. A minimum diameter of 2μm and maximum diameter 20 um were achieved in these experiments. It is believed that the maximum feasible diameter of channels that can be fabricated by this technique was not determined by the experiments. It is also found that laser power, gas flow rate and traveling speed are correlated and need further study. The maximum length of smooth channel was found to be 5 mm ~ 10 mm which is long enough for some microfluidic channel applications.
Appl. Y. Du. 12(10): p. Korn. and P. Optics express. H. Rapidly Prototyped Three-Dimensional Nanofluidic Channel Networks in Glass Substrates. 2005.. Mitsuyu. R. 64(23): p. 175 2. Dou. A . 5083-5088. Phys. Genin. 1855-1859. Chem. Photowritten optical waveguides in various galsses with ultrashort pulse laser. Sands. Fabrication of high-aspect ratio. 79(3): p. 26(2): p. Y. K. Sugimoto. and K. Writing waveguides in glass with a femtosecond laser. A. 2120-2129. Schaffer. Bado. D. Lett. Hirao. M. 2005. 1994. K. A. and F. Miura. An. D.. Ke. N. and E. 6. Garcia. Bellouard. Phys. 1996. 2000. Liquid-assisted femtosecond laser drilling of straight and three-dimensional microchannels in glass. Mazur. 1997. 2001. J. and G. Phys. 3329-3331. C. 7. Lett. & Processing. 71(6): p. and C. Gong. Choi. Y. 77(16): p. Machining of transparent materials using an IR and UV nanosecond pulsed laser. 5. Mourou. 13(6): p. H. and Q. Hwang. 9. Appl.Mater. 21(21): p.. Processing. micro-fluidic channels and tunnels using femtosecond laser pulses and chemical etching. Anal. Yang. and K. Grigoropoulos. 2004.3. Laser-induced breakdown by impact ionization in SiO2 with pulse widths from 7ns to 150fs. A Matl. Squier. Hirao. Lett. 2004. Miura. 601-608. J. 71(23): p. K. and A. T. 8. Phys. Inouye. J. 3. Liu. Dugan. Optics express. Hunt. Appl. 1729-1731. 93-95. Micromachining bulk glass by use of femtosecond laser pulses with nanojoule energy. K. 3071-3073. Sci. Hasselbrink.Salleo. Brodeur. Lett. T. 4. T. Li. Sci. Qiu. Simultaneous multi-microhole drilling of soda-lime glass by water-assisted ablation with femtosecond laser pulses. 605-612. Opt. X. . Said. A.6 Reference 1. Opt. Davis. G. E.4. Appl.
Schaffer. Richardson. 3rd ed. 13.10. 525-546. in Applied Physics. Phys. J. Microeng. 2005. T. D. Schoen. A. C. Femtosecond laser micro-structuring and refractive index modification applied to laser and photonic devices. Biomaterials science. C. and D. T. Micromech. SPIE. San Diego: Academic press. Lippert. 663-666.C. Rivero.. 12. 201106. Zoubir. Schaffer. 15. O2/Ar based discharges. B. Englewood Clifs: Prentice Hall. Plasma process polym. Lopez. 1996. C. M.. Hoffmann. 2005. Interaction of femtosecond laser pulses with transparent materials. Groisman. Femtosecond laser-drilled capillary integrated into a microfluidic device. Kleinfeld. Petit. 86(20): p.. 15347-4. and J. 2001. C. Appl. 2004. V: p. Kim. C. 11. C. Yang. Transport Processes and Unit Operations. A. 14: p. An Introduction to Materials in Medicine. 176 . Deep reactive ion etching of commercial PMMA in O2/CHF3. Ratner. Richardson. A. 1993. Interaction of photons with polymers: From surface modification to ablation. 2: p. 14. and L. Ding. Zhang. K. Harvard University: Cambridge. 2003. L. Lemons. and C. F. 16. Geankoplis. Lett.
The ideal scaffold requires appropriate structures such as various sizes of porosity. migration. The ultimate goal of femtosecond laser ablation is to preserve “neotissue” environment for cell growth after 177 . laser ablation is being investigated for producing micro-scale structures on ES-PCL fiber meshes.femtosecond laser ablation was used to produce various structures on ES-PCL tissue scaffold and it was used for cell growth. In a previous article . and differentiation and a degradation rate that closely matches regeneration rate of the desired natural tissue . electrospun (ES) nanofiber mesh of poly-ε caprolactone (PCL) [1-4] is widely used as a tissue scaffold to grow cells and to proliferate biological tissues [5-9]. To satisfy these requirements.1 Introduction Poly-caprolactone (PCL) is a Food and Drug Administration approved biodegradable polyester that has been used for many biomedical applications .CHAPTER 5 MEASUREMENT OF FEMTOSECOND LASER SCATTERING IN ELECTROSPUN POLYCAPROLACTONE NANOFIBER 5. In tissue engineering. growth. sufficient surface area and a variety of surface chemistries for cell adhesion.
the sizes and locations are much more random than is usually considered in the analysis of photonic crystals. It is noted that absorption and scattering of optical radiation within an ES polymer bears similarity to other previously published analyses for different applications. and has potential to control surface topography to produce specific microenvironments that can influence cellular growth patterns.6μm). it was concluded that absorption and scattering of radiation within the fiber mesh was significant and that multiple scattering needed to be further analyzed for a predictive understanding of the ablation process and structures produced by it. It was also noted that the interaction of the laser radiation with ES fiber material was quite different from that usually observed for solid materials. Direct-write femtosecond laser ablation process can provide a flexible and rapid structuring of tissue scaffold. and high peak irradiance for high precision machining as well as non-clean process. Photonic crystals consisting of periodically-spaced near wavelength-sized structural units have been widely studied recently. The knowledge of the wavelength dependence of scattering is also useful for the determination of structural features of tissue. Femtosecond laser with a central wavelength of 775 nm and pulse width of 150 fs. The width of channels ablated by a scanned focus spot (10μm ~ 50μm) was dramatically greater than the focus spot diameter (~ 1. Based on the results. was considered to be a promising process for ES fiber structuring because of its non-thermal interaction. 10]. Some biological tissues have a fibrous structures and their interaction with light has been studied by many researchers [11-16] because it provides much useful information for treatments . The diameters of the ES PCL fibers used in this work are also on the order of the laser wavelength (775nm). However.structuring the tissue scaffold [3. 178 .
Furthermore. The reflectance and transmittance of the sample are relatively easily measured. 179 . during laser ablation of ES polymer meshes. in the cases of interest for this work.and backward-directed radiation is much more divergent than the incident beam so measurement of the corresponding power is best done by an integrating sphere to ensures capture of all radiation [13. In cases where scattering is an important component of the material interaction with a material sample. so the absorption of light is exponentially decreases as thick thickness of sample increases. when light irradiates a sample of a material. the dimensions of the sample are large relative to the fiber spacing so incident rays of light are scattered multiple time until they are completely absorbed by material or escape from the sample. Also. 5. it may be reflected. and/or scattered inside and outside of the samples. interaction with the material is non-linear . but differentiation of absorption and scattering inside of the sample is more difficult. 15.1.As shown in Fig. 16]. absorbed. the forward.
For convenience. isotropic and optically linear materials irradiated by an plane wave of infinite extent. Scattering of light from a single particle in a surrounding medium depends on the size of particle and wavelength of light in the medium.1 Light propagation inside and outside of a fiber mesh supported by a transparent substrate. Analytical solutions for scattering from a single cylinder of infinite length are based on the summations of terms involving functions ' ' J n ( y ) J n ( x ) − mJ n ( y ) J n ( x ) tan α n = ' ' J n ( y ) N n ( x ) − mJ n ( y ) N n ( x ) (5. The reader is referred to the cited works for details.2) and 180 . Mie theory is applicable for single scattering where wavelength and particle size are comparable[18-20] and are generally valid for homogeneous. a number of the relevant results from analysis of scattering are briefly summarized below.Figure 5.
Jn is Bessel function of the first kind of order n.2 Definition of TE and TM modes (E: electric field. n1 is refractive index of the cylinder material and n0 is index of the surrounding material. M: magnetic field.2. 5.3) where αn and βn are phase angle for TE and TM polarization respectively as defined in Fig. and Nn is the second Hankel function of order n. k: wave vector) The extinction efficiencies for single scattering of radiation with TE and TM polarizations can be calculated as 181 .4) and y= 2πa ⎛ n1 ⎞ ⎜ ⎟ λ ⎜ n0 ⎟ ⎝ ⎠ (5. Dimensionless coordinate positions x and y in (5.5) where a is the radius of the cylinder.3) are defined as x= 2πa λ (5.2) and (5.' ' mJ n ( y ) J n ( x ) − J n ( y ) J n ( x ) tan β n = ' ' mJ n ( y ) N n ( x ) − J n ( y ) N n ( x ) (5. Figure 5.
6) QTM −ext where (5. An early solution that is based on a two-flux assumption (radiation is separated into forward.10)  μ E dI = −( E a + E s ) I + s dz 4π ∫ π p ( Ω. Ω’: all direction. It was originally developed to describe 182 . (5. I: intensity.7) an = and tan α n . Ω ) I ( Ω ) d Ω ' ' 4 ' (5. μ: the cosine of the angle θ between direction Ω and the surface normal of the sample. tan α n − i (5.9) Analysis of multiple scattering requires solutions based on the radiative transfer equation as described in Eq. p: phase function.QTE −ext 2 ∞ = ∑ Re( bn ) x n = −∞ 2 ∞ = ∑ Re( an ) x n =−∞ (5.10) where.8) bn = tan β n tan β n − i (5.and backward-directed components) within a scattering medium is sometimes referred to a Kubelka-Munk theory.
the back reflection from an infinitely-thick sample of material R∞ is expressed as . Assuming the case of isotropic scattering. the assumption of diffuse irradiation does not allows analysis of scattering from a sample illuminated with collimated laser light . matched boundaries.11) where S is the K-M absorption coefficient and K is the K-M scattering coefficient. 25] 183 . Mudgett et al derived a 22 flux approximation and that related the two-flux KM coefficients.multiple scattering from paint films but also applies in other applications [22-28]. (5.12) The calculations can be as follows [23. to physical absorption (μa) and scattering coefficients (μs) described in Eq. and diffuse irradiation at the sample surface.12). K and S. Even though the two-flux theory is a simple and a popular method for analyzing the optical properties of scattering layers. R∞ ≈ 1 + K − S K2 S 2 +2 K S (5. I ( z) = I 0e −( μa + μs ) z (5.
2Experiment setup 184 . Burger et al. for diffused irradiation. a0 and a1 are phase function coefficients. procedures of integrating sphere measurements and analysis to estimate absorption and scattering coefficients. generalized Kubelka-Munk theory using three flux approximation and they came up with an identical results with Mudgett et al. and widened ablation spot diameter which is believed to be caused by multiple scattering mechanism of light inside ES PCL fibers. In this article. determination of ablation threshold for the fiber material.14) For analysis of laser ablation.27 μs S (5.13) where. The following relationship can be made for collimated or directional illumination  μa K ≈ 0. 29.μa K 2 ≈ S (3a 0 − a1 ) / 4 μ s (5. 30]. 5. for isotropic material a0 =1 and a1 = 0 Three-flux approximation using the equation of radiative transfer was used to calculate the scattering and absorption coefficients [23. differentiation between scattering coefficient and absorption coefficient is important and is the subject of the analysis described in this paper.
The mechanical and electrical properties of PCL such as dielectric values and mechanical viscosity.000. 185 . However. NJ) and pumped through a 20 gauge stainless steel blunt tip needle at a flow rate of 24 ml/hr under an applied voltage of 24 kV. St. The sizes of deposited fibers are randomly distributed between 200nm and 2μm. The volume ratio which is defined as the ratio of fibers to a unit volume is estimated to be 20% and the distribution of fiber diameters are counted and the results are shown in Fig. Fiber was deposited onto flat glass substrate coated with a thin (~150nm) conductive film (ITO. The thicknesses of electro spun fibers varied between 30 μm and 130 μm.3. 5. Phillipsburg. fibers which diameters are bigger than 1μm tend to have irregular diameters which are caused by instability of stream of liquid-state fiber jet. Louis. Indium Oxide Tin Oxide). The density of the resulting mesh and other properties of the polymers are summarized in Table 1. Aldrich. affect the electrospinning process so the ES fiber diameter and fiber density cab vary. MO) dissolved in acetone (Mallinckroff Chemicals.PCL fibers were electrospun from a 12 wt% solution of PCL (Mw 65.
The measurements of the transmittance and reflectance were done with an integrating sphere technique. The linear polarization axis was normal to the scan direction and a half-wave plate optic was used to change the polarization. pulse duration of 150 fs. and reflectance measurements. The beam had a diameter of 5 mm.3 and was directly applied on the sample without going through focusing optics. The transmitted or reflected lights were collected by an integrating sphere (IS236A.3 Bulk density g/cm3 1.07* Heat Capacity kJ/(kg K) 1. An average power of 1. an M2 quality parameter of 1.2 58 ~ 63 Table 5.Figure5. Temp.24 Solid density g/cm4 1.3 Electro spun fibers (left) and fiber diameter distribution (right) Melting temp.6watt (6kHz) was attenuated to several mW with a series of thinfilm polarizing beam splitters (PBS) and wave plates. transmittance. 186 . CPA-2110) was used as a collimated light source which was used to fabricate a tissue scaffold structuring.6watt (2kHz) and 2. The integrating sphere is a general purpose sphere for many applications like laser power. Thorlabs. and a repetition frequency of 2 kHz / 3kHz / 6kHz. Tm (C) Glass trans.1 Material properties of PCL Femtosecond laser (Clark-MXR. The laser was amplified by a Ti:Al2O3 regenerative amplifier operating at a central wavelength of 775 nm. Tg (C) −65 ~ −60 Thermal conductivity W/(m K) 0.14 ES mesh density g/cm3 0.
) that can measure a wide wavelength range (300nm ~ 1100nm) and both anode and cathode grounded polarities.4. collecting the transmitted or reflected light flux from a thin tissue sample. 187 . has been specially arranged in order to prevent direct irradiation of the detector even under large divergent input beams .Tetrafluoro Ethylene (PTFE) based high reflective bulk material. The port for the direct connection of photodiodes (SM05PD . being independent of the direction. 5. It has a diameter of 2” and has an inner surface covered by highly reflecting Poly. NJ).25GHz sampling rate and provides the relative intensity of transmitted or reflected lights.Newton. and thus allows it to be probed by a small photo diode at one single position. The scattering by the integrating sphere transfers the directed transmitted and reflected light to diffuse light fluxes. The photodiode converts the light intensity to current and trans-impedance converts current signal to voltage signal. Thorlabs Inc. Oscilloscope (TDS3012B. The detail system setup is depicted in Fig. Tektronix) is used to record the signal at 1.
Figure 5. To see the saturation of signal. the output signal decreases as the thickness 188 . To measure the maximum transmittance (100%).4 System setup for transmittance and reflectance measurements As shown in Fig. NJ) and calibrated with the output voltage from integrating sphere. To calibrate the sensor. the incoming laser beam was measured with a calibrated power meter (PM120. Thorlabs. Newton.5. laser beam was applied without any samples and the output signal was measured. With a sample. laser beam was increased up to 20mW and input power and output sensor signal was linearly correlated and no saturation was observed. The sensor was calibrated to convert the transmittance.5. samples were mounted in the inlet or the outlet port of the integrating sphere for transmittance and reflectance measurement of samples. respectively.
with a reflecting mirror. The ambient light was also turned off for zero calibration at 0% transmittance. the output port was open so that all laser light exits to ambient which replicates 100% of absorption or 0% reflectance.5. To measure the 0% of reflectance. with cap and without cap.of sample increases. provided us the scale of reflectance. the laser beam needed to be re-directed to the sample. To have 100% of reflectance calibration. Fig.5 (b). the output port was blocked with a cap with same material as bulk material coating. For reflectance measurement purpose. Figure 5. The difference of two signals.5 System setup for transmittance and reflectance measurements 189 .
5. but the measurement was not dependent on laser beam polarization. The slope was very linear over the entire range of incident power and the slope represents the transmittance of sample at the give thickness. input power of samples 40um 50um 130um 30um 65um 90um "no sample" 190 . The polarization was rotated by 90 degrees.6. the input power and output power of integrating sphere was compared with different thickness of PCL samples. 9 8 Output Power (mW) 7 6 5 4 3 2 1 0 0 2 4 6 8 10 Input Power (mW) Figure 5. The incident power was varied from 0.5.7mW to 8mW by using a half wave plate rotation and PBS.6 Output power vs.3Results As shown in Fig.
it is assumed that the transmittance is 100% at zero thickness. the slope of transmittance decay.Since the thickness varies within the same sample. However. the extinction coefficients (μt = μa + μs) were estimated as 0. The thickness measurement of samples was done with micrometer.5. Figure 5. respectively. The transmittance was plotted as shown in Fig. The solid PCL is prepared by melting the fiber meshed PCL.7. the zero thickness assumption is still a pending question. In our analysis. and variation was observed to be +/-10μm which includes measuring force variation and sample thickness variation within the sample.7 Transmittance measurement for fiber PCL (circles) and solid PCL (squares) 191 . the experiments were done multiple times at different locations in the same sample. The transmittance decreases exponentially and the result was compared to that of solid PCL’s.0013 μm-1 and 0. By curve fitting.0085 μm-1 for solid PCL and fiber PCL.
11) and (5.15) where. μa is absorption coefficient. (5.5.13). μs is scattering coefficient 100% 80% Reflectance (%) 60% 40% 20% 0% 0 100 200 Thickness (μm) 300 Figure 5. The other relationship can be estimated from the experimental measurement of transmittance of samples.As the same manner. The reflectance is asymptotically converging to 65% as thickness of sample increases up to 300μm. and the ratio of KubelkaMunk absorption and scattering coefficients can be calculated from Eqs. μt = μ s + μ a (5. This value can be considered as R∞ .8 Reflectance measurement 192 . the reflectance of PCL fiber mesh was measured as shown in Fig.8.
we found that 97% of scattering from the absorbed light inside the material.17) It is noted that F0 = 2 Fav .16) where w0 is Gaussian beam radius and F0 is maximum (or peak) fluence. 2 F (r ) = F0 exp − 2r 2 w0 ( ) (5. the ablation threshold fluence. related to pulse energy Ep as 2 F0 = 2 E p / πw0 (5.By using above described approaches. 25. the diameter D of the area near the center of the beam from which material is removed is then 193 . we analyzed the absorption and scattering coefficients. we have about 98% of scattering from the absorbed light inside the material. The fundamentals of femtosecond laser ablation have been summarized in previous sections and there are many references elsewhere . The total extinction coefficient which combines absorption and scattering coefficients can be estimated by curve fitting of transmittance using exponential function fitting. 29]. With 3 flux theory approximation which assumes a collimated light illumination. Kubelka-Munk theory. If F0 > Fth . and three flux approaximation. we can also separate the absorption and scattering coefficients from the total extinction coefficient [22. The ablation threshold for fiber meshed PCL was measured. By using the total extinction coefficient from the experiment. For our purposes we can consider single pulse removal of a material with threshold ablation fluence Fth by a beam having a Gaussian radial fluence profile. With 22 flux theory which assumes a diffused light illumination.
9 Ablation diameters vs.9. Based on Eqs.18) Fth can be found by curve fitting a semi-logarithmic plot of D2 measured for various F0 and calculating the intercept at zero fluence. we calculated the ablation threshold as Fth=0.10 0. Fth=1 J/cm2 .2 D 2 = 2w0 ln (F0 / Fth ) . 1000 900 800 Diameter2 (μm 2) 700 600 500 400 300 200 100 0 0.15 0. and they are plotted as Fig.25 Fluence(J/cm2) Figure 5. We measured the width of ablation and fluence. It is believed that the multiple scattering including backscattering inside ES PCL fiber widens 194 .5.00 0. This number is lower when it is compared to bulk PCL ablation threshold. (5. Fluence One of reasons of lower ablation threshold is due to multiple scattering.18).06 J/cm2. (5.16) – (5.20 0.05 0.
5.6μm) was applied on the surface of PCL fiber mesh. initial focused spot diameter (d0=1.the affected area of laser beam irradiation which in turn decreases laser fluence or ablation threshold calculation. As we can see in Fig. The experimental results support this assumpition and we observed that the width of ablation was as big as 8 μm to 15 μm even lowest laser energy we applied (Fig.11) Figure 5. but multiple scattering occurs which redirect incident laser beams and scattered laser beams.10 Increased ablation diameter by multiple scattering 195 .10.5.
Based on multiscattering theories. The plasma expands rapidly shortly after the femtosecond laser pulse and is observed to persist for many 10’s of nanoseconds [34.4Conclusions and future works In this section. the femtosecond laser scattering inside ES fiber is presented. 5.11 Ablated fiber with femtosecond laser (E=500nJ) It is also observed that there are two distinct slopes in the graph which does not follow logarithmic curve fitting in Fig. Integrating sphere was used to collect the transmitted and reflected light from samples.Figure 5.5. One of reason is that a strong plasma radiation may affect the widening the width of ablation at higher fluence region which are visually noticeable. about 97% and 98% of light absorbed inside material 196 .9. 35].
The ablation threshold is measured to be 0. It is believed that the multiscattering and backscattering of light contributes to lowering ablation threshold and widening ablation width.scatters with 22 flux and 3 flux assumptions. 1.6μm. respectively. A computer simulation based on Maxwell equations can help us understand the multiple scattering phenomena inside ES PCL fiber and that remains as future work. The measured minimum ablation width is 8μm which is wider than theoretically calculated diffraction limited focus spot size. 197 .094 J/cm2 which is lower than that of bulk material 1J/cm2.
Lee. 26(37-46). and J. Inai. J. Yang. 8. Baumgarten.5. Biomaterials. Laurencin. 11. produced by electrospinning. and W. R. 2004.and microfiber meshes for tissue-engineering matrix and scaffold based on newly devised multilayering and mixing electrospinning techniques. and D. Ma. Hu. Li. Khil. Mesoscopic spatial designs of nano. S. K. Y. D. 2007. C. Electrospinning of silk fibroin nanofibers and its effect on the adhesion and spreading of normal human keratinocytes and fibroblasts in vitro. T. Polymer. Min.. J. J. Nanotechnology. and I. Xu. Ra. 1971. Colloid Interface Sci. I. Electrospinning for tissue engineering scaffolds. D. Nam. G. Nam. 25(1289-1297). Baum. Johnson. P. 44(26): p. Matsuda. T. Shin. Y. and J. C. 36: p. Farson. 24: p. Lannutti. Nanometre diameter fibres of polymer. 198 5. 2003. 613-621. 25: p. Biomaterials.R. 2004. Tomasko. M. 2003.E. W. Lee. and D. 216-223. 1287-1294. Caterson. Material science and engineering C.. P. 2007. S. 27: p. M. Chun. and S. Structuring electrospun polycaprolactone nanofiber tissue scaffolds by femtosecond laser ablation. 7. Kim. 1996. Biomaterials. R. 504-509. Characterization of nano-structured poly( e-caprolactone) nonwoven mats via electrospinning. 10. 6. Kidoaki. 2002. H. Journal of Laser Applications. 71-79. Y. B. Lee. Kim. Farson. . Kattawar. Aligned biodegradable nanotibrous structure: a potential scaffold for blood vessel engineering. Biomaterials. Mishchenko. Ko.through far-infrared spectral region. Biomedical Mat. Lee. A biodegradable nanofiber scaffold by electrospinning and its potential for bone tissue engineering. 2. 2005. J. Ramakrishna. Kotaki. H. G. Huang. Park. 3. Fu. Applied Optics. Choi. H. Wei. H. Tuan. D. 7: p. B. 9. and T. 60: p. and Q. Yoshimoto. Res. Lannutti. 44: p. and F. Electrospun nanofibrous structure: A novel scaffold for tissue engineering. 2077-2082.. Kwon. 5512-5523. Y. H. M. Reneker. 4. H. Electrostatic spinning of acrylic microfibres. D. 877-886. Terai. Scattering and absorption property database for nonspherical ice particles in the near. 2005. Vacanti.5References 1.
2001. Hoel. J. techn. 1997. S. and I. 15. A. and S. Phys. 469-473. Unified framework for homogeneous and inhomogeneous optical media. Kubelka. 2001. 42(16): p. Beitrage zur Optik truber Medien. Fabbri. Boyer. and F. Frederick. Applied Optics. G. IS&T's NIP 17: Int. Applied Optics. Franceschini. Light scattering by small particles. 297-328. and J. Opt. 19. He. and J. 1960. Soc. Bosch. DeWeerd. 2004. J. Lai. Mourad. T. ed. 44(10): p. Munk. Am. 36(4): p. C. Marley. 1999. Li. J. A. P. A. 70(6): p. 330: p. J. Phys. Applied Optics. Simon. 18. 593-601. Hersch. and R. Zijp. 17.. Bigio.12. G. 21(10): p. 14. 2001: p. Chandrasekhar. 377-445. L. T. Conf. Wang. A. 1908. Applied Optics. J. 1931. Molenaar. 1845-1849. 2002. Niklasson. 34: p. Applied Optics. 38(10): p. T. New York: Dover Publications. J. Cox. Gaffney. Radiative Transfer. 1942-1952. Blazer. Miklavcic. 1957. N. J. Experimental measurement of the refractive index of biological tissues by total internal reflection. M. H. Johnson. . 2003. on digital printing technologies. Ann. and S. 2nd. 2005. P. Emmel. and J. Fuselier. Physik. S. 21. Aerosol Science and Technology. 535-549. 3063-3072. M. Preditions and measurements of scattering and absorption over broad wavelength ranges in tissue phantoms. P. 2068-2077. speziell kolloidaler Metallosungen. New York: Dover Publications. Fantini. R. Linden. Kruse. Characterization of spatial and temporal variations in the optical properties of tissuelike media with diffuse reflectance imaging. Am. Determination of Kubelka-Munk scattering and absorption coefficients by diffuse illumination. 20. 16. Optical characterization method for black pigments applied to solar-selective absorbing paints. F. J. C. E. K. 199 13. An experiment to measure Mie and Rayleigh total scattering cross sections. Mie. Extending Kubelka-Munk's theory with lateral light scattering. An empirical method for the determination of the complex refractive index of sizefractionated atmospheric aerosols for radiative transfer calculations. Ein Beitrag zur Optik der Farbanstriche. and A. Mourant. B. 24. Orel. 40(10): p. Tesfamichael.C. 620-625. 949-957. Drayton. 1672-1681. and Z. Yang. Wackelgard. Z. 23. 12: p. Gunde. van de Hulst. Revised Kubelka-Munk theory II. 22. Baird. Zeits f.
25. 309-317.. 30. 1977. 1999. P. G. Sci. S. 31: p. Ultrashort pulse laser interaction with dielectrics and polymers. and W. 2: p. Caps. T. 1759-1762. www. S. B. 525-546. Phys.. Interaction of photons with polymers: From surface modification to ablation. V. Femtosecond and picosecond laser microablation: ablation efficiency and laser microplasma expansion.com. 21662185. 28. A . 729732. Appl. Kuhn. Khun. Pivovarov. 29. J. Levy. 34. 2005. 10(7): p. A review of the optical properties of biological tissues. T. P. 1996. 1990. 35. H. E.. Caparative evaluation of absorption coefficients of KCL:Eu2+ and CaF2:EU2+ using a spectrophotometer and an integrating sphere. and A. and B. Diffuse reflectance and transmittance spectroscopy for the quantitative determination of scattering and absorption coefficients in quantitative power analysis. Ebel. Meynadier. Ploss. A. 26(12): p. Mudgett. Patterson. 200 . Advances in polymer science. Sardar. P. Shelden. Applied Optics. and A. Kilmentov. Lippert. S381-S383. Welch. Dausinger. D. The role of plasma in ablation of materials by ultrashort laser pulses. IEEE journal of quantum electronic. 1485-1502. 378-382. Processing. Petite. Spectrosco. Prokhorov. W. Quant. Burger. 33.. 2004. 2001. O. Appl. and J. Richards. 16(3): p. S. Appl. Kruger. Gobert. J. S. Kautek. Semerok. Thorlabs. T. C. and L. Konov. 31. Inc. Kubelka-Munk optical properties of a barium sulfate white reflectance standard. 1323-1329.thorlabs. R.. 79(3): p. Multiple scattering calculations for technology. Perdrix. and J. 69: p. 51: p. 168: p. Quantitative determination of the scattering and absorption coefficients from diffuse reflectance and transmittance measurements applied to pharmaceutical powders. 51: p. D. 1997. and F. Fricke. Burger. Cheong. Stockton. T. 27. Fricke. accessed 7/11/07 J. and L. Kononenko. Breitling. Salle. Prahl. J. M. Applied Spectrsc. 26. Garnov.Mater. Applied Optics. Plasma process polym. 1997. 32. Phys. 1971. Electronics. 247-289.
Whitesides. Du. Annu. Mourou. 1706-1716.” Synthetic Metals 138. W. Li. C. and G. and G. 1997.References 1. X. 3. 1992. M. 375-377. Journal of Chromatography. Res. Key engineering materials. Bado. and F. Laser ablation and micromachining with ultrashort laser pulses. Said. Du. Machining Oxide Thine Films with an AFM: Pattern and Object Formation on the Nanometer Scale. W.E. Electrospun nanofibrous structure: A novel scaffold for tissue engineering. Su. 257(5068): p. Li. C. 1997. 60: p. 4. Z. 800-804. Science. 10. 6. Laurencin. 613-621. and C. X. Y. M. A(1038): p. Gao. 33(10): p. 2006. R. Electrospun nanofibrous structure: A novel scaffold for tissue engineering. Rev. D. Dugan. micro-fluidic channels and tunnels using femtosecond laser pulses and chemical etching. Sci. IEEE journal of quantum electronics. Zheng. Y. D. Lieber. 7. Mourou. 2120-2129.. 60: p. Laser ablation and micromachining with ultrashort laser pulses. 2004. Kim. J. and F. Res. Hu. Tak. 2004. R. Chen. 33(10): p. 9. 613-621. 12(10): p. Three dimensional micromachining based on AFM. Biomedical Mat. and X. Fabrication of high-aspect ratio. Liu. Kim. R. S. 315: p. Optics express. and G. Tuan. 8.S. 28: p. Tuan.T. Zhang. Z. and G. 33(10): p. IEEE journal of quantum electronics. Liu. Du. 1997. 1706-1716. Biomedical Mat. Y. 153-184. A. C. D. Ko. “Novel patterning method using Nd : YAG and Nd : YVO4 lasers for organic light emitting diodes. 239-245.. Caterson. Lin. 2002. Kim. and Y. Y. Vacuum assisted thermal bonding of plastic capillary electrophoresis microchip imprinted with stainless steel template.N. 2002. 11. 201 . 5.E.. M. Soft lithography.H. K. 1998. Laurencin.H. Bellouard. IEEE journal of quantum electronics. Lee and S. X. Kim.Xie.B. Kim. and P. Y. 2. Caterson. Ko. Liu. Laser ablation and micromachining with ultrashort laser pulses. Mater. 1706-1716. Mourou. Xia. 497-500 (2003).
40(10): p. 22. Ph. T. 13. Niklasson. Proceedings of symposium H on photonic processing of surfaces . 525-546. 453 . K. Prahl. Groisman. www. Hermann. Gunde. R. 2004. W. Burger. 453 . Cheong. 201106. Technical Glass Products. Phys. 21. Itina... Inc.12. Delaporte. Lippert. 181-187. Applied Optics. and M. A. Proceedings of symposium H on photonic processing of surfaces . and Z. Modeling of metal ablation induced by ultrashort laser pulses.454. Fang. Rajeev. IEEE journal of quantum electronic. 525-546. Lippert. Hoel. M. and C. and D. and J.C. Kim. 525-546. 2000.454. Lippert.thorlabs. and A. Ph. T. T. 25. J. Wackelgard. 2: p. Bhardwaj. 1672-1681. Nanotechnology. 20. Strong field ionization inside transparent solids. Phys. 21662185. Opt. E-MRS: p. Fricke. A. T. Plasma process polym. Hermann. S. 1990. Optical characterization method for black pigments applied to solar-selective absorbing paints. Interaction of photons with polymers: From surface modification to ablation. 51: p. 2: p. B: At. 11: p.. Sentis. T. Machining characterization of the nanolithography process using AFM. Sentis. Caps. 14. Applied Spectrsc. Plasma process polym. Schaffer. Phys. E. Mol. Properties of fused-quartz. 23. D. J. 2001. 1997. 2: p. 202 . and J. A review of the optical properties of biological tissues. Kleinfeld. 16. Tesfamichael. C. Itina. 17. Appl. 2005. 86(20): p. Microslide specification. J. 309-317.. P. Khun.. 24. Modeling of metal ablation induced by ultrashort laser pulses. 39: p. 2005. 18. T. T. Quantitative determination of the scattering and absorption coefficients from diffuse reflectance and transmittance measurements applied to pharmaceutical powders. Interaction of photons with polymers: From surface modification to ablation.. 2007. accessed 7/11/07 15.. V. T. Plasma process polym. Interaction of photons with polymers: From surface modification to ablation. Lett.. S397-S407. Femtosecond laserdrilled capillary integrated into a microfluidic device. P.com. 2007. Thermofisher scientific. Rayner.Thin films and devices. 2005. 2005. G. Weng. Corkum.Thin films and devices. and M. 19. Welch. Thorlabs. Orel. 2004. J. T. Chang. 2006. E-MRS: p. Delaporte. 26(12): p.
I. A. G. Fallnich. Momma. Simon.V. 2716-2722. A. 2001: p. Conf. S. Phys. Extending Kubelka-Munk's theory with lateral light scattering. Applied surface science. Saleh. Ebel. Garnov. 35. S. J.V. V. and J. Soc. Audouard. 31(5): p. Garnov. “The role of plasma in ablation of materials by ultrashort laser pulses. 27. D. Fricke. Momma. N. 28. 30. Am. 36. Stuke. Opt. S. 14(10): p. Fortunier. Tunnermann. H. T. J. 76: p. The role of plasma in ablation of materials by ultrashort laser pulses. H. C. K. Ploss. Jacobs. 1323-1329. Breitling. H. Welling. 14(10): p.M. S... Nolte.. Nolte. S. Anal. Momma. S. on digital printing technologies. T. Emmel. 203 . F. B.M. Lai.A.26. B. Nolte. 4-6. and F. Kamlage. Kuhn. S. and H. Alvenslebe. Welling. Pivovarov. P. 247: p. S. Kononenko. 563-567. Ablation of metals by ultrashort laser pulses. and M. Chichkov. C. B. and Teich. Diffuse reflectance and transmittance spectroscopy for the quantitative determination of scattering and absorption coefficients in quantitative power analysis.J. S. and H. P. Tunnermann. Ablation of Polytetrafluoroethylene (Teflon) With Femtosecond UV Excimer Laser Pulses. Hersch. 238-242. Femtosecond UV excimer laser ablation. C. E. Prokhorov. 37. Ostendort. 29. Klimentov. D. 32. 2004. 68(6): p. Am. Ablation of metals by ultrashort laser pulses. Chem. 34. 1999. Dausinger. S. 31. Appl. Welling. and H. and R. Fundamentals of photonics. and M. 1175-1183. Kononenko. Mourad. 44: p. Lett. Chichkov.” Quant. B. 1987. and R. 2D Calculations of the thermal effects due to femtosecond laser-metal interaction. Quantum Electronics. 2716-2722. Breitling and F. Wellegehausen. Pivovarov. Cao. 469-473. 33. S. and L.materials science & processing. Harzic. Kuper. Dausinger. J. Wellegehausen. 1997. Kuper. S. A. P. 199-204. Prokhorov. 1997. Konov. X. 2005. Appl. Appl. A. Lee. Soc. 51: p. 378-382 (2001). Spectrosco. Phys. R. 31. Huot. 54: p. Applied physics A . Elec. Konov. 1989. Opt. S. Klimentov. Valette. Jacobs. 378-382. B. Burger. 2001. IS&T's NIP 17: Int. T. Stuke. 1999: Wiley. V. A. 1997. B: Photophysics Laser Chemistry. A Packaging techinque for polymer microfluidic platforms. C. B. Polarization effects in ultrashort-pulse laser drilling.
Chen. Kwon. Biomaterials. S.A12 45. 43. Tankiewicz. Pivovarov. Rosenfeld and EEB Campbell. Biomaterials. Applied Optics.” J. 204 . Bosch. Kidoaki. Mesoscopic spatial designs of nano. P. D. Phys. S. 44. 13(6): p. Prokhorov. S. I. Ultrafast laser micromachining and patterning of thermal spray multilayers for thermopile fabrication. Appl. 1999. 14: p. 47. Takayama. 2001. Garnov. I. 2363-2376. 46. Patterning proteins and cells using soft lithography. T. 506-513. 2000. V. P. App. Simultaneous multi-microhole drilling of soda-lime glass by water-assisted ablation with femtosecond laser pulses. Prokhorov. Quant. Bonse. 13167. Chandrasekhar. An. A. Baudach. 42. 39. R. Ashkenasi. and J. Dausinger. Determination of Kubelka-Munk scattering and absorption coefficients by diffuse illumination. 2005. and G. S. Breitling. The role of plasma in ablation of materials by ultrashort laser pulses. Matsuda. Li. Surface Sci. Electronics. S. 2005. Zijp. Y. Kononenko. 2001. Kilmentov. 20: p. Mesoscopic spatial designs of nano. Kilmentov. and F. E. Kautek. Radiative Transfer. Matsuda. A. 26(37-46). 2068-2077. and F. Ohya and Yasutaka Takahashi. 26(37-46).and microfiber meshes for tissue-engineering matrix and scaffold based on newly devised multilayering and mixing electrospinning techniques. Gong. D. J. R. Garnov. R. 48. S. J. D. Whitesides. Ultrashort pulse laser ablation of polycarbonate and polymethylmethacrylate. Bel Hadj Tahar. Microeng. Kidoaki. 83. J. S. Micromech. 40. D. New York: Dover Publications. 41. S. T. 555-560. 1960. Ban. Quant. T. 2631-2645 (1998). V. 154: p. H.38.13173 (2000). 1855-1859. B 62. Konov. A. 31: p. Longtin.” Phys. “Tin doped indium oxide thin films: Electrical properties. The role of plasma in ablation of materials by ultrashort laser pulses. and T. 1999. J. Y. S. Konov. Yang. Q. 378-382. Molenaar. Kane. Electronics. Kononenko. Rev. Y. Dou. Kruger. Sampath. J. Ostuni. Dausinger. and W. R.and microfiber meshes for tissue-engineering matrix and scaffold based on newly devised multilayering and mixing electrospinning techniques. 2004. Ingber. 2005. 31: p. and T. Stoian. and R. and Q.. Pivovarov. “Coulomb explosion in ultrashort pulsed laser ablation of Al2O3. S. 378-382. Optics express. Kwon. R. Breitling. 49. Biomaterials. S. Gambino. 38(10): p.
Kattawar. Femtosecond pulsed laser ablation and deposition of thin films of polytetrafluoroethylene.. The effect of damage accumulation behabiour on ablation thresholds and damage morphology in ultrafast laser micro-machining of common metals in air. Colloid Interface Sci. Frederick. Yang. P. Appl. 71-79. Bulgakova. 12: p. and F. J. and L. 44(26): p. Baum. Applied surface science. 2430-2441. P. techn. Phys. 36: p. “Effect of substrate absorption on the efficiency of laser patterning of indium tin oxide thin films. Soc..M. Multiple scattering calculations for technology. Aerosol Science and Technology. 42074212 (1999) 58.through far-infrared spectral region. Mudgett. 1971. P. E. Appl. 2004. Takai. P. 54. A. Kubelka. Surface Sci. M. 2004. and T.. and T. Baird. and extinction by thin fibers. G. 52. 1931. Hu. 51. Colloid Interface Sci. Coyne. 34: p. 57.” J. Richards. 59. Wei. M. 56. 275-287. J. 1971. P. M. P. Baumgarten. Zeits f. 1999. Applied surface science. Yavas and M. Waterman. J. 2004(221). Applied Optics. Mishchenko. Munk. Marley. E. Fu. J. Y. Mannion. Blazer. Drayton. Am. 223: p. 2005.. B. J. G. Ein Beitrag zur Optik der Farbanstriche. 2001. Physical Review E. Magee. 36: p. 2005. 71-79. . 1971. 22(11): p. The effect of damage accumulation behabiour on ablation thresholds and damage morphology in ultrafast laser micro-machining of common metals in air.50. and P. H. Vendan. Mannion. J. Electrostatic spining of acrylic microfibres.R3500. Scattering. 55. R3498 . Applied Optics. C. 85. Womack. Gaffney. Huang. 535-549. Coyne. Molian. Baumgarten. absorption. P. Glynn. 5512-5523. 60. Electrostatic spinning of acrylic microfibres. P. An empirical method for the determination of the complex refractive index of sizefractionated atmospheric aerosols for radiative transfer calculations. J. and J. Opt. G. Physik. Scattering and absorption property database for nonspherical ice particles in the near. P. 275-287. and Q. 223: p. 593-601. N. Glynn. 53. N. H. 60(4): p. O'Connor. Possibility of rarefaction shock wave under short pulse laser ablation of solids. 1485-1502. O'Connor. 205 . 10(7): p.
1997. Cheng. Rivero. Spielmann. B. Sasnett. 62. Farsari. C. 2000. Bercier. Influence of master fabrication techniques on the characteristics of embossed microfluidic channels. G. Femtosecond Optical Breakdown in Dielectrics. and P. 4584-4589 (2004). and F. Thorsen. Unified framework for homogeneous and inhomogeneous optical media. M. Zoppel.” Appl. 132-142. W. Scherer.. Revised Kubelka-Munk theory II. S. J. 2005. 1989. 2004. Schulzgen. and H. 21(10): p. 15: p. Mourou. and Fotakis. Lenzner. Lopez. Filippidis. SPIE. 65. S. A. Science. Reider.K. M.. Krausz. Z. Soc. 68. Zoppel. Microeng. L. 15: p. Journal of Laser Applications. Kapur. M. J. Molian. and V. N. Genova. 69. L. 43. Giridhar. and L. M. Lab on a Chip. Rossier. Seong.F. 70. P. 1942-1952. 113-116. Anal. and S. 1786-1789. Richardson. Farsari. “Femtosecond pulsed laser micromachining of glass substrates with application to microfluidic devices. G. K. UV laser machined polymer substrates for the development of microdiagnostic systems. Elbandrawy and M. Micromech. Irizarry. G. 4076-4079.. and P. P. Sartania. 1998. 67. C. G. Girault. Physical Review Letters. Unger. 15347-4. Microeng. M. G. 17861789. M. A. Monolithic microfabricated valves and pumps by multilayer soft lithography. Efficient femtosecond laser micromachining of bulk 3C-SiC. Kautek. M. 18-28. J. 64. J. New York: Adam Hilger. A. 2003. Reider. 10(1): p. A. J. 2003. Opt. 121-127.. “Optical characteristics of femtosecond laser micromachined periodic structures in Si <100>. 72. 1998. Hall. ed. 66. Petit. Propagation of multimode laser beam-the M2 factor. and S. 7137-7143 (2006). A review of ultrashort pulsed laser ablation of materials. S. 206 . Miklavcic. 2005.61. Femtosecond laser micro-structuring and refractive index modification applied to laser and photonic devices. Jackson. S. V: p. 2035-2042. and Fotakis. Yang. Opt. Richardson. Roberts. Micromech. Quake. 3(2): p. Efficient femtosecond laser micromachining of bulk 3C-SiC. M. Kruse. Khulbe. H. Opt. 63. Peyghambarian and M. M. 69: p. 45.” Appl. Esch. 288: p. Chem. 71. T. G. Zoubir. M. D. Mansuripur. Shirk. M. Gupta. Am.C. 80(18): p. C.
S. 2003. Lan. K. M. 159-164. M. K. 71(23): p. Ra. “High energy femtosecond Yb cubicon fiber amplifier. Richardson. 1287-1294. 775-779. 75. 44: p. Daele. 2002. T. 2003. Lett. H. 3329-3331. Flexible fabrication of microlenses in polymer layers with excimer laser ablation. Soviet Physics JETP. Femtosecond laser-induced damage and filamentary propagation in fused silica. Express 13. 44: p. Naessens. Khil. M. Lamouroux. 207 . L. Kim. 2005. Hong. and A. and D. M. 26: p. G. Polymer. 74. Rapidly Prototyped Three-Dimensional Nanofluidic Channel Networks in Glass Substrates. L. M. E. J. H. and L. J. Applied surface science. 763-769. Goh. 1965. B. Y. Biomaterials. J. Sudrie. Mitsuyu. 151-164.. L.E. Franco.T. 89(18): p. J. G. Femtosecond laser deep hole drilling of silicate glasses in air. Wang. K. Khil. Hunt. and A. Mysyrowicz. K. Hasselbrink. H. Lee. 84. P. Richardson. L.. Anal. Kim. Ionization in the eld of a strong electromagnetic wave. Z.C. Tech (B). Physical Review Letters. T. K. Chem. B. Ke. Miura. K.” Opt. Lee. 208209: p. Z. 76. 1287-1294. Microeng. 186601. and M. 1307-1314. 83. Cho and M. S. Hirao. 1997. 2001. Appl. Lim. Teoh. Tiaw. Y. Ra. Lee. L. 183: p. 2003..73. 2002..A. Vac. Qiu. Characterization of nano-structured poly( e-caprolactone) nonwoven mats via electrospinning. Hartl .. and R. 79. Lee. Imeshev. 80. I. 82.C. H. Sci. and K. Esashi. Photowritten optical waveguides in various galsses with ultrashort pulse laser. Applied surface science. and K. 77. Laser surface modification of poly(epsilon-caprolactone) (PCL) membrane for tissue engineering applications. and S. 2005. 77(16): p. 2003. 5083-5088. Femtosecond laser ablation of thin films for the fabrication of binary photomasks. S. Phys. Polymer. Venkatakrishnan.. P. Fermann. Shah. Li.L. and D. 12: p. Shah. 20: p. 2545-2549. Baets. 21(6): p. K. Inouye. Keldysh. Characterization of nano-structured poly( e-caprolactone) nonwoven mats via electrospinning. Micromech. Prade. 81. 4717-4722 (2005). Liu.. Smooth surface glass etching by deep reactive ion etching with SF6 and Xe gases.O. Tzortzakis. B. 78. A.
27: p. Hong. 92.. Tomasko. and P. Park. J. V. 1542 . Michel. Hurd. 86. 1997. Chiu. 2000. McDonald. 1996. 21: p. Electrophoresis. Applied Optics. A. Fabrication of microfluidic systems in PDMS. Chiu. J.I. Polymer embossing tools for rapid prototyping of plastic microfluidic devices.. Johnson. 2000. D. 2006. 87. 504-509. 94. 89. Anderson. Polymerization optimization of SU-8 photoresist and its applications in microfluidic systems and MEMS. 88(183113): p. McDonald. Carnal. D. 2007. 1-3.P. McDonald. D. Limbach. Electrospinning for tissue engineering scaffolds. J. Fabrication of microfluidic systems in PDMS. Buckley. J. 91. 27-40. T. 27-40.R. 21: p. G. Vollet. “Femtosecond fiber lasers with pulse energies above 10 nJ.85. J. 90. N.. J. Morita. and D. Whitesides. Wu. 96. and I.D. Narashimhan. D.and microfluidic channels. 2006. Zhang. D. G.W. Lett. 21(21): p. Anderson. Micromech. Diamond tip cantilever for micr/nano machining based on AFM. 2004. Farson. Reymond. and F. Micromech. J. and K. 11: p. 2005. J. Takano. 45: p. 93.Reneker. D. V. Tan. J. Boyer. Lett. 2001. a. Appl. Mistry. Schueller. Sosnowski. 1888-1890 (2005). Writing waveguides in glass with a femtosecond laser. 949-957. Wise. Electrophoresis. J. Hirao. Schueller. F. Lagger. T. and H. Whitesides. Miura. T. 505: p. Bigio. H. 2(3): p. and S. Phys. Davis. Mourant. International Journal of Machine Tools & Manufacture.. 1729-1731. 36(4): p. Yang. 2002. N. Lab on a Chip. L. 30. Microeng. and G. Girault. Material science and engineering C. Lett. O. J. Ray.. J. Enhanced machining of micron-scale features in microchip molding masters by CNC milling. 88. Materials Science Forum.1550. Sugimoto. H.L. Ma. O. J. Gobry.O. K. K. and N. 95. Rossier. Preditions and measurements of scattering and absorption over broad wavelength ranges in tissue phantoms. 20-26. Mecomber. J. Plasma etched polymer microelectrochemical systems. Fuselier. Yalisove. C. 96-103. J. H. Wu. Microeng. Gong.” Opt. Lannutti. P. and G. Ilday and T. 145-150. Opt. 208 . 14: p. D.D. 79-84. K. F. J. D. Femtosecond pulsed laser direct write production of nano. K.
282-286. Appl. Material science and engineering C. 2005. 15: p. 105. Ihlemann. 2nd. Advances in polymer science. H.” Adv Polymer Sci 168. 2005. A biodegradable nanofiber scaffold by electrospinning and its potential for bone tissue engineering.B. Kautek. Advances in polymer science. Chuang. Phys.. 63(20): p. 247-289. Kuznia. 24: p.K. 2004. Oslen. 44(10): p.. Biomaterials. He. Experimental measurement of the refractive index of biological tissues by total internal reflection. Lett. 1147-1156. 2077-2082. J. 2002. Shin. “Ultrashort Pulse Laser Interaction with Dielectrics and Polymers. H. D. Applied Optics. H. Klank. J. Lai. Khan. 247-289. 98. 104. 99. 1993. Wang. Andideh. C. and O. and W. 2077-2082. 1996. J. J. and D. New York: Dover Publications. 1996. van de Hulst. 2003.R. H. M. Lannutti. Excimer laser micro machining of inorganic dielectrics. Li. Vacanti.W. J. Kruger. 2003. H. Geschke. A. Ultrashort pulse laser interaction with dielectrics and polymers. 430-438. 107. Applied surface science. Electrospinning for tissue engineering scaffolds. Kruger. and J. 242-246.-R. Farson. Applied surface science. Kruger. E. Kautek.M. 100. Ma..W. CO2 laser micomachining and back end processing for rapid production of PMMA based microfluidic systems. 247–289 (2004). Micromech. Biomaterials. Lab on a chip. Cheng..K. J. Vacanti. Terai. 504-509. 96(98): p. D. Shin. 168: p. Microeng. 168: p. J. Femtosecond-pulse visible laser processing of transparent materials. 27: p. D. A biodegradable nanofiber scaffold by electrospinning and its potential for bone tissue engineering. 108. 2004. Yoshimoto. Young. 1957. 297-328. and Y. 102. Reactive ion etching of GaN in silicon tetrachloride plasmas. 209 . and J. Tomasko. 1845-1849. M. Crack-free direct writing on glass using a low power UV laser in the manufacture of a microfluidic chip. Ultrashort pulse laser interaction with dielectrics and polymers.. ed. Adesida. 109. Light scattering by small particles. 101. Krüger and W.97.. and A. H. Y. and T. Yoshimoto. J. 106: p. T. Terai. a. Y. C. J.Y. and W. 2(4): p.C. Kautek.. 24: p. J. 103. and J. 2007. Wei. 2777-2779. Z. a. 106.
Basting. J. Dehghani. Excimer laser: A new tool for precision micromachining. Arai. Structuring electrospun polycaprolactone nanofiber tissue scaffolds by femtosecond laser ablation. 2005. 116. Biol. R. Huang. Z. S297-S301. X. J. 14: p. and D. and J. Jiang. Chou. Huang. J. 210 . 27(12): p. 115. Pogue. 1319-1335. and Y. Smart Materials and Structurs. Applied Optics. Bovatsek. Applied Optics. The effects of internal refractive index variation in near-infrared optical tomography: a finite element modeling approach.4.J. Brooksby. Ingber. Soft Lithography. D. Direct-write patterning of ITO film by high pulse repetition rate femtosecond laser ablation. H. 27(12): p. and D. H. 2005. 2007.6Ti49. 2000. R. H. Experimental investigations machiniability of Ni50. Lett. 122. Lannutti.. Sensors and Actuators.. 3: p. 118. Chen. E. J. H. 121. Johnson. 113. and D. G. 335-373. Wu. 2007. of Biomedical Engineering. Structuring electrospun polycaprolactone nanofiber tissue scaffolds by femtosecond laser ablation. H. H. Lannutti.Yu.Z. and Y.. Vishwanath. 119. Choi. Whitesides. G. Arai.. Choi. Zheng.F. Liu. 112. Appl. K. S297-S301. Basting. D. Becker. 1995. Endert. Phys. Bovatsek. H. Austin. 130-135. H. Annu. 2002. Takayama. J. and D. Patzel. 14: p. 114. Cao. H. 120. Choi. and D. and S. 111. Nam.4. Nam. J. Excimer laser: A new tool for precision micromachining. Journal of Laser Applications. 83: p. 2007. Paulse. D. H. B.. 2007. 2003. Direct-write patterning of ITO films by high pulse repetition frequency femtosecond laser ablation. 48: p. 1995. A. Farson. J. Journal of Laser Applications. Ashkenasi. J.6Ti49. 174-176. Soft lithography in biology and biochemistry.. Farson. Rev. Endert. Phys.. Choi. S. 1998. H. D. R.O. J. Smart Materials and Structurs. Farson.110. 117. 81(1): p. Fabrication of 10nm enclosed nanofluidic channels. and D. WTEC hyper-librarian. Heim. and U. W. A. H. 1319-1335. B. 2713-2727. Experimental investigations machiniability of Ni50. Med. Liu. E. Tegenfeldt. Hot embossing as a method for the fabrication of polymer high aspect ratio structures. Optical and Quantum electronics. Optical and Quantum electronics. Whitesides.P. Ashkenasi. 2001. Wang. H.
K. Mainfray. F. 2001. 127. Phys.a. 2003.Atomic molecular and optical physics. 4005 . Ann. A. 11: p. 9(3): p.C. Stockton. Multiphoton ionization of atoms. 86(12): p. and W. 133. J. 27-32. Microeng. Huang. Wokaun. G.L. IEEE journal of quantum electronic. Tikhonchuk. 3063-3072.M. Shelden. C. Fantini. M. and B. J. Physical Review Letters. E. 1991. T. Microeng. Ionization of atoms in the tunneling regime with experimental .R.L. 131. Gobrecht. 132. Thin solid films. Laville. 2001. and V.. 1543-1548. S. and C... 1999. Applied Optics. 54(10): p. Kubelka-Munk optical properties of a barium sulfate white reflectance standard. 25(19): p. Micromachined pre-focused 1XN flow switches for continuous sample injection. J. J.D. 377-445. 453: p.4020.J. O. 567-573. Vidal. M. A novel micro-machining method for the fabrication of thick-film SU-8 embedded micro-channels.. C. Luther-Davies. 2004. 5049 . Choong. B. Watt. 124. Lee. 1977. Physical Review A. D. Kopitkovas. Gamaly. and S. 330: p. Micromech. Fabbri. Hwei.evidence using Hg atoms.123. 5(6): p. Ablation of solids by femtosecond lasers: Ablation mechnasim and ablation thresholds for metals and dielectrics. E. 134. and M. Mie. Micromech. Piper. B. T. 2002.B. Tay. F. Physics of plasma. 211 . J. Applied Optics. C. 1908. Hung. Gibson. Critical-point phase separation in laser ablation of conductors. 125. and B.S. Chaker. Enhanced polymer ablation rates using high-repetition-rate ultraviolet lasers. Journal of Physics B . A. 16(3): p. Illy. LeDegoff. Characterization of spatial and temporal variations in the optical properties of tissuelike media with diffuse reflectance imaging.1372. F. M. T. Tunneling ionization in the multiphoton regime.. Ke. 41(9): p. Patterson. F. and J. 1333 . G. 949-957. 2001. 1992. Ilkov.C. 42(16): p. 3135. Margot. Barthelemy. and J. G.. 1990. G.. 729732. Franceschini. Surface micromachining of UV transparent materials. Sabsabi. Beitrage zur Optik truber Medien. F. Withford. Reports on progress in physics.5052. G. B. 129. 128. David. 126. 11: p. 2573 -2576.. G. Rhodes. a. E. speziell kolloidaler Metallosungen. 130.
Nanometre diameter fibres of polymer.135. R. Rosenfeld. and L. 605-612. D. 150: p.L. Nanotechnology. 18(3): p. 210-215. and C. D. Lee.. D. Experimental and theoretical anlysis of direct-write laser micromachining of polymethyl methacrylate by CO2 laser ablation. D. Korn. S. M. “Modeling optical breakdown in dielectrics during ultrafast laser processing. D. R. & Processing. 223-228 (2003). Current. Caparative evaluation of absorption coefficients of KCL:Eu2+ and CaF2:EU2+ using a spectrophotometer and an integrating sphere. Sci. 212 . 1996. Femtosecond laser bulk micromachining of microfluid channels in poly(methylmethacrylate). 1996. 79(3): p. I. Choi. 81(10): p. 1759-1762. Journal of Applied Physics. J. Laser induced breakdown by impact ionization in SiO2 with pulse widths from 7ns to 150fs. and M. Phys. 216-223. J.. Poate. J. I.M. CPA 2110 User manual. Phys. 140. Muller. X. C. Stoian. A Matl. Melngailis. 3124-3131 (2001)..P. Rosenfeld.E. D. D. Lizuain. Appl. Longtin. 142. Tasch. 2006. D. Liu. J. C. 6513-6561. Surface Sci. Hwang. Journal of Laser Applications. T. M. Surface damage threshold and structuring of dielectrics using femtosecond laser pulses: the role of incubation. Ion beams in silicon processing and characterization. 79(3): p. C. D. Lu. Rubia. 138. 44(5): p. 136. Phys. Sci. A.. Liquid-assisted femtosecond laser drilling of straight and three-dimensional microchannels in glass. Ashkenasi. 024901-1:6. N. Mourou. O. Holland. Appl. H.. 141. Chanson. Appl. Farson. 139. Processing. 1999. 101-106. Borland. Gomez. and I. Law. 64(23): p. Optical engineering. 77. J. J. Yuan. “Fundamentals and advantages of ultrafast microstructuring of transparent materials. and L. 2004. 1997. Squier. Phys. 1994. Phys. Clark-MXR. D. Fan.B. Levy. 144. and A.. Campbell. 2005. Choi. 2 ed. A-Mater. 3071-3073. 40. Appl. Magee. E. Reneker. 137. Hertel. Das. 051105:1-8. 145. G. Stoian. Lett. 143. and A. M.H.” Appl.V. 146. I. T.” Appl.. 2004: Clark-MXR. Chun. Femtosecond laser ablation for microfluidics. G. Opt. 2007(101): p. Sardar. Grigoropoulos. Appl. Eaglesham. and S.P. and G. Du.G. Bulgakova and E. 7: p. Ozaita. produced by electrospinning. Ashkenasi. J.
877-886. MA. O2/Ar based discharges. 2001. Biomaterials. Johnson. Geankoplis. C. International Journal of Machine Tools & Manufacture. Harvard University: Cambridge. Micromech. Mazur. 945-949. 93-95. 151. Technol. in Physics. C. Ocana. Molpeceres .147. Schaffer. 2007. C. Zhang. Brodeur and E. Opt. S. 25: p.J. 158. P. Transport Processes and Unit Operations. 2004. Inai. 14: p. Lett. Urbina and J. Microeng.1794. 15. 154. and S. Lu. 2001. C. Kotaki. Xu.” J. Kovacevic. Modeling of waterjet guided laser grooving of silicon. 26: p. Englewood Clifs: Prentice Hall. 1993. and S. 43: p. Anthony. Biomaterials. Focused ion beam microfabrication in Foturan(TM) photosensitive glass. Brodeur. 149. Micromachining bulk glass by use of femtosecond laser pulses with nanojoule energy. Li. Mao. Laser induced breakdown and damage in bulk transparent materials induced by tightly focused femtosecond laser pulses. 2005. C. 213 . J. Interaction of femtosecond laser pulses with transparent materials. 26(2): p. Interaction of femtosecond laser pulses with transparent materials. M. 3rd ed. and K. Fabrication of optical waveguides in thermosetting polymers using hot embossing. 2004. Y. Jiang. P. Schaffer. A. Choi. Deep reactive ion etching of commercial PMMA in O2/CHF3. 925 . in Applied Physics. 150. 1784 . Ramakrishna. A. Aligned biodegradable nanotibrous structure: a potential scaffold for blood vessel engineering. Garcia. 663-666. C. 12: p. C. J. Chen. Ding. 157. 14: p. C. R. Meas. J. 2001. 115-119.936. L. J. Mazur. C. 152. Microeng. Docker. p. 2004. Carabe. 148. Direct micro-patterning of biodegradable polymers using ultraviolet and femtosecond lasers.L. 17: p. J. Schaffer. 173. C. Micromech. and D. Micromech. S. Micromech. 7642-7649. Lauzurica . 2003. Prewett. J. C. Yang. 2001. Sci. D. Gandia. Microeng. C. Aguilar. 155. Schaffer. Microeng.. “Microprocessing of ITO and a-Si thin films using ns laser sources. C. 156. Harvard University: Cambridge. and R. 1271-1278 (2005). and E. 153.
M. B. B. Seddon. F. Alvensleben. B. Petite and A. Lerberghe. Laser Micro/Nanoengineering. Hoffmann.81. Ratner. Salle. 65: p. An Introduction to Materials in Medicine. B. Simon. B. Phys. Koinkar. and D. S381-S383 (1999). Nolte. 63(2): p. Fisette and M. Journal of microelectromechanical systems. Nanosecond to femtosecond laser-induced breakdown in dielectrics. Biomaterials science. “Three-dimensional microfabrication inside photosensitive glasses by femtosecond laser. 36. 74(12): p. Gobert. Ultrafast dynamics of nonequilibrium electrons in metals under femtosecond laser irradiation. Rethfeld. 1653-1655. 25(1289-1297). P. Physical Review B. Stuart. 76 . Applied Physics A: Materails Science & Processing. Kaiser. and H. Park. B. Shao. G. Nam. Laser-induced damage in dielectrics with nanosecond to subpicosecond pulses. Lett. A. 1994. Femtosecond and picosecond laser microablation: ablation efficiency and laser microplasma expansion. A. Tunnermann. Fost. M. M. Electrochimica Acta. B. 166. B. Physical Review B. Meynadier. Perry. Phys. Stuart. 64(13): p. San Diego: Academic press. Herman. The application of excimer-laser micromachining for the fabrication of disc microelectrodes. and A.” Appl. T. Lemons. 1. Physical Review Letters. J. O. 169. 7066-7072 (1997). Appl. 2000. 167. and M. Motsegood. Girault. 783-791. and S. 164. L. Beebe. Feit. “Indium tin oxide overlayered waveguides for sensor applications. 163. G. and J. 1996. J. 39(6): p. Min. picosecond and nanosecond laser ablation of solids. Semerok. Meunier. Shore. 2004. Luff. A. 1994. A – Mat. Three-dimensional Microchannel fabrication in PDMS elastomer.” Appl. Vicanek. 2002. S. Perdrix. Momma. 162.. Processing 69. Biomaterials. 1749-1761. 1996.214303-11.159. B. Jo.S. Lee. M. 170. 214 . 1995. 9(1): p. F. 1996. 165. Nanoindentation hardness measurements using AFM. Electrospinning of silk fibroin nanofibers and its effect on the adhesion and spreading of normal human keratinocytes and fibroblasts in vitro. Femtosecond . Schoen. Feit. Opt. Kim. B. and V. 160. Y. C. S. 109-115. B. 7-11 (2006).J. Sci. Lee. 161. 214303-1 . 168. Rubenchik. B. and W. Chichkov. Perrone. K. 53(4): p. B. Y. Wilkinson and G. 2248-2252. and G. Bhushan.” JLMN-J.
Optics and Lasers in Engineering. S104-S111. 176. Rosengarten.. Ben-Yakar. 15: p. Byer. A. Simon. Y. Ashkenasi. J. Femtosecond laser ablation properties of borosilicate glass. 2002. 175. Appl. Fabrication of high-aspect-ratio microstructures using excimer laser. 96: p. Phys. Sci. 215 . 71(6): p. Illy. 1447 1459. 5316 . and R. 601-608. Appl. C. Phys. 172.5323. Piper. 827-847.): p. Cher.171. 2004. G. 2004. 41: p. M. Tseng. J.. and D. A . 173. An experiment to measure Mie and Rayleigh total scattering cross sections. Genin. Holland. and J. Struct. 177. and K. Plasma etching of thermoset polymer films for microchannels. J. 69(Suppl. Am. and F. A. Smart Mater. 830-836. DeWeerd. E. Sands. 1999. Phys. Ultrashort laser pulse damage threshold of transparent materials and the role of incubation.Mater. Processing. R. Glover. and D. Lorenz. Rosenfeld. Ben-Yakar. Ma. Appl. 40: p. Linden. Phys. Harkin. D. Thermal and fluid processes of a thin melt zone during femtosecond laser ablation of glass : the formation of rims by single laser pulses. S373-S376. Stoian. Byer.. Salleo. High speed UV micro machining of polymers with frequency doubled copper vaport lasers. 620-625. Stone. T. 70(6): p. and J. A. Machining of transparent materials using an IR and UV nanosecond pulsed laser. A. 2006. Chen. 2007. Cox. 178. J. Phys. Appl. 1995. 2000. 1(3): p. 174. R. Ashmore. and H. IEEE journal of quantum electronic.
This action might not be possible to undo. Are you sure you want to continue?
We've moved you to where you read on your other device.
Get the full title to continue listening from where you left off, or restart the preview.