Comparative Study of the Capabilities of Various Micromachining Processes

Contents

1 Introduction
1.1 Background and Motivation
1.2 Research Objective, Scope, and Research Plan
1.3 Thesis Content

2 Laser Machining
2.1 Introduction
2.2 Laser Parameters
2.3 CO2 Lasers
2.3.1 Introduction
2.3.2 Capabilities
2.3.3 Industrial Applications
2.3.4 Costs
2.4 ND:YAG Lasers
2.4.1 Introduction
2.4.2 Capabilities
2.4.3 Industrial applications
2.4.4 Costs
2.5 Excimer Lasers
2.5.1 Introduction
2.5.2 Techniques
2.5.3 Capabilities
2.5.4 Industrial Applications
2.5.5 Costs
2.6 Femtosecond Lasers
2.6.1 Introduction
2.6.2 Ablation process
2.6.3 Techniques
2.6.4 Capabilities
2.6.5 Industrial Applications
2.6.6 Costs
2.7 Conclusion

3 Mechanical Machining
3.1 Introduction
3.2 Tool Materials
3.2.1 Diamond
3.2.2 Tool Steel and Tungsten Carbide
3.3 Microdrilling
3.3.1 Drilling Machines
3.3.2 Drills
3.3.3 Cutting Process
3.3.4 Capabilities
3.3.5 Industrial Applications
3.3.6 Costs
3.4 Micro-Turning
3.4.1 Turning Machines
3.4.2 Turning Tools
3.4.3 Difficulties of Microturning
3.4.4 Cutting Process
3.4.5 Development of a micro-lathe
3.4.6 Capabilities
3.4.7 Industrial Applications
3.4.8 Costs
3.5 Micromilling
3.5.1 Milling Machines
3.5.2 Milling Tools
3.5.3 Cutting Process
3.5.4 Development of a meso-scale milling machine
3.5.5 Capabilities
3.5.6 Industrial Applications
3.5.7 Costs
3.6 Conclusion

4 Microforming
4.1 Introduction
4.2 Microforming Techniques and Capabilities
4.2.1 Extrusion
4.2.2 Cold Forging
4.2.3 Deep Drawing
4.2.4 Embossing/Coining
4.2.5 Air Bending
4.2.6 Blanking and Punching
4.3 Forming Process
4.3.1 Material Behavior
4.3.2 Friction
4.4 Machines
4.5 Forming Tools
4.6 Costs
4.7 Conclusion

5 Micro Electro-Discharge Machining (Micro-EDM)
5.1 Introduction
5.2 Principles of Material Removal in EDM
5.3 Factors influencing the Micro-EDM Process
5.4 Micro-EDM Techniques
5.4.1 Micro Electro-Discharge Die-Sinking
5.4.2 Micro Electro-Discharge Drilling
5.4.3 Micro Wire Electro-Discharge Machining
5.4.4 Micro Wire Electro-Discharge Grinding
5.4.5 Micro Electro-Discharge Grinding
5.4.6 Micro Electro-Discharge Milling
5.4.7 Uniform Wear Method
5.4.8 Summary of the Machining Techniques
5.5 Capabilities
5.5.1 Micro Electro-Discharge Die-Sinking
5.5.2 Micro Electro-Discharge Drilling
5.5.3 Micro Wire Electro-Discharge Machining
5.5.4 Micro Wire Electro-Discharge Grinding
5.5.5 Micro Electro-Discharge Grinding
5.5.6 Micro Electro-Discharge Milling
5.5.7 Uniform Wear Method
5.6 Machines
5.7 Industrial Applications
5.8 Costs
5.9 Conclusion

6 Conclusion
6.1 Summary and Comparison
6.2 Process Selection
6.3 Research and Application Issues

List of References

A Surface Roughness Measures

B Machine Specifications
B.1 CO2 Laser
B.2 Nd:YAG Laser
B.3 Excimer Laser
B.4 Femtosecond Laser
B.5 Drilling Machine (Via Drilling)
B.6 Turning Machine
B.7 Milling Machine
B.8 Forming Machine
B.9 Micro Electro-Discharge Machine

List of Tables

1.1 Methods used for micromachining[1]

2.1 Comparison of direct writing method with mask projection method[2]

2.2 Excimer gas mixtures[3]

2.3 Energy per photon at different excimer wavelengths[4]

2.4 Ablation rate using a KrF excimer laser[5]

2.5 Comparison of different beam techniques[6]

3.1 Wear of Diamond by Sliding Metals at a sliding speed of 100-200 m/sec[7]. .

3.2 Materials that can be diamond machined[8]

3.3 Materials that can be diamond machined via ductile mode/Diamond grind- ing[8]

3.4 Typical Diamond Turning Parameters[8]

3.5 Recommended machining conditions depending on workpiece material[9]

3.6 Maximal feedrates for good surface qualities[8]

3.7 Microturning Parameters and Results

3.8 Parameter variation in micromilling using FIB-tools[10]

3.9 Comparison of different results of mechanical micromachining

4.1 Deep drawing of 100 µm thick sheets[11]

5.1 Summary of EDM techniques and machinable shapes[12]

5.2 Material removal rates in micro-EDM

5.3 Comparison of different micro-EDM techniques and their capabilities

6.1 Comparison of micromachining techniques

6.2 Advantages and disadvantages of laser micromachining processes

6.3 Advantages and disadvantages of mechanical micromachining processes

6.4 Advantages and disadvantages of microforming and micro-EDM processes

B.1 CO2 laser specifications

B.2 Nd:YAG laser specifications

B.3 Excimer laser specifications

B.4 Femtosecond laser specifications

B.5 Drilling machine specifications

B.6 Turning machine specifications

B.7 Milling machine specifications

B.8 Forming machine specifications

B.9 Micro electro-discharge drilling/milling machine specifications

List of Figures

1.1 Size range, absolute accuracy, and relative accuracy in precision machining[13].

2.1 Overview of typical laser characteristics for Nd:YAG, excimer, and titanium:sapphire lasers[14]

2.2 Laser beam interaction and material removal

2.3 Material removal rate per pulse in stainless steel depending on the sample thickness[15]

2.4 Material removal rate as a function of sample thickness for different wave- lengths[16]

2.5 Applications of Nd:YAG lasers, depending on laser parameters[15]

2.6 Entrance side (31 µm) and exit side (26 µm) of a percussion-drilled hole in 1.46 mm stainless steel (10W, 10kHz, 355nm)[17]

2.7 Setup of LAM66 and an example structure, machined in Wc/Co with a depth of 150 µm[18]

2.8 Static mask projection at the example of a hole mask (left) for the simulta- neous production of a hole array (right)[19]

2.9 Linear ramps produced in PET at 248 nm, using the moving mask technique[20]

2.10 Moving workpiece: a) General principle[19] ; b) Pyramids machined using the moving workpiece technique[21]

2.11 HAM - Expected section of the cavity machined with a mask with holes of two different diameters[22]

2.12 PTFE, processed at 193 nm (a) and 157 nm (b). Fused silica processed at 193 nm (c) and 157 nm (d).[23]

2.13 Relationship between drilling rate of silicon and laser fluence at different beam diameters. Dots are experimental values and lines are calculated[24]

2.14 10-µm-diameter hole in an alumina ceramic plate 0.5 mm thick (aspect ratio of 50) drilled with a KrF laser[25]

2.15 Holes drilled with a twist drill bit (a) and a KrF laser[21]

2.16 Feedhorn structure machined in polycarbonate using an excimer laser[26]. .

2.17 KrF laser-drilled ink jet printer nozzles in polyimide[21]

2.18 Comparison of the undesired side effects associated with long laser pulses (a) and ultrashort laser pulses (b)[27]

2.19 Machining of holes smaller than the laser spot size, using the Gaussian distri- bution[28]

2.20 Squared diameters D[2] of the surface damage region in dependence of the laser Fluence F0 for TiN by number of pulses (N=1 and N=100)[29]

2.21 Diameter D of the surface damage region versus the number of applied laser pulses for TiN in different polarization states[29]

2.22 Crater depth and ablated Volume versus the number of applied laser pulses for TiN. F0 = 0.43J/cm[2] ; 130f s)[29]

2.23 Influence of the ambient pressure using the mask projection technique[30]. .

2.24 Comparison of a trepanned (a) and a percussion drilled hole (b)[31]

2.25 Influence of the polarization on the drilling quality of high-aspect ratio holes[30].

2.26 Schematic illustration of SNOM technique[32]

2.27 Holes drilled with Ti:Sapphire laser (120 femtoseconds) a) in air b) in vac- uum, compared to c) a hole machined by a ND:YAG laser (λ=1.06 µm; nanoseconds)[33]

2.28 0.6 µm hole in aluminium, machined with 130 fs pulse[34]

2.29 Hole machined into stainless steel with diameter 35 µm and a depth of 0 mm[28]

2.30 Hole through single crystal diamond sample using femtosecond laser (a) and Nd:YAG laser (b)[31]

2.31 Comparison of different lasers material processing techniques

3.1 V-block arrangement for microdrilling concentricity[35]

3.2 Working principle of a specialized ultra-precision drilling machine[36]

3.3 Geometry of microspade drills[35]

3.4 Peck drilling cycle used to clear holes from chips[35]

3.5 Drill point geometry and nomenclature[37]

3.6 Reliability functions using conventional microdrilling, R1(t), and vibrating drilling R(t)[38]

3.7 Force components of main cutting edge[39]

3.8 Shear plane model of the second cutting edge[39]

3.9 Drilling system (a), drill geometry (b), and manufactured drill (c) for drilling small holes[40]

3.10 Microhole with a diameter of 6.7 µm and a depth of 10 µm (feedrate: 0 µm/s[40]

3.11 Hub of copper gear made with photosensitive polyimide process[41]

3.12 Hole in brass with a new tool (a), after the 6 holes have been drilled (b), and the 4th hole in stainless steel (c)[42]

3.13 Micro chuck of a micro lathe turning system[43]

3.14 Diamond single point tool[43]

3.15 Deflection of workpiece under cutting force[43]

3.16 Machining strategies, microturning[44]

3.17 Workpiece tensions for different machining strategies[44]

3.18 Reasons for machining errors in microturning[44]

3.19 Factors influencing the surface roughness in ultraprecision diamond turning[45].

3.20 Sharp-edge cutting model (a) and round-edge cutting model (b)[46]

3.21 Forces in the round-edge cutting model[46]

3.22 Cutting force (a) and thrust force (b) per unit width in RECM and SECM[46]

3.23 Examples of shape cutting[43]

3.24 Test workpiece for NC-lathes with micrometer dimensions[47]

3.25 Micromilling tools shaped by focused ion beam sputtering[5]

3.26 Flycutting tool (a) and two endmills (b and c)[48]

3.27 250 µm milling tool[49]

3.28 Cutting edge workpiece interaction in micromilling[50]

3.29 Bending of the tool if no chip is formed until the ninth pass[51]

3.30 Burr formation in micromilling[49]

3.31 Definition and shape of burrs in micromilling[52]

3.32 Built-up edge at a milling tool and cutting tool marks in aluminium[49]

3.33 Simulated micromilling operation of a ductile iron workpiece[53]

3.34 Meso-scale milling machine[54]

3.35 NOH-mask in glass[55] (rough cut [a]/finished [b]) and in copper [c][56]

3.36 Geometry of a pseudo ball end mill[55]

3.37 2-fluted milling tool made of HSS[57]

3.38 2 mm diameter micromilled pattern (a) with steps (b) and thin, straight wall (c)[57]

3.39 Walls (a) and steps (b) in nickel silver, created using a diamond end mill[44]

3.40 Wedges in nickel silver (a) and sinusoid structures in brass (b), created using diamond tools[44]

3.41 Method (a) and result (b) of milling V-shaped grooves[58]

3.42 Groove machined in aluminium[10]

4.1 Schematic illustration and photograph of a miniature backward extrusion ma- chine[59]

4.2 Extruded gear with a 10 µm module (a) and with a 20 µm module (b;c)[59].

4.3 Effect of specimen size and grain structure on the shape of the extruded part[60]

4.4 Extruded structure with a diameter of 0.5 mm and a wall thickness of µm[60]

4.5 Deep drawing by local laser heating[61]

4.6 Bulging height, embossing depth, and punch force in aluminium[60]

4.7 Schematic illustration of the LADI process[62]

4.8 Quartz mould (a) and imprinted structure in silicon (b)[62]

4.9 Springs produced using microbending[60]

4.10 Entrance (a) and exit (b) of a punched hole in brass[63]

4.11 Effect of grain size on bending forces and yield strength[60]

4.12 Effect of grain size on microformability[64]

4.13 Principle of open and closed lubricant pockets[60]

4.14 Cup for electron gun made from molybdenum[60]

4.15 Die produced with micro wire-EDM[60]

5.1 Setup of an EDM system[65]

5.2 Material removal process in EDM and micro-EDM[66]

5.3 Surface modification, due to the discharge process[65]

5.4 Flushing techniques using a jet of dielectric fluid[65]

5.5 Flushing techniques using a relative movement between electrode and work- piece[65]

5.6 Principle of micro electro-discharge die-sinking[48]

5.7 Triangular hole without (a) and with (b) jumping of the tool electrode[67]

5.8 Principle of micro-WEDM[68]

5.9 Taper in micro-WEDM[65]

5.10 Principle of WEDG[66]

5.11 Machinable shapes by WEDG[69]

5.12 Principle of micro-EDG[68]

5.13 Electro-discharge milling - profiling (a), pocketing (b), and grooving (c)[1]

5.14 Different types of electrode wear[66]

5.15 Layer-by-layer machining[70]

5.16 Tool paths in uniform wear method[70]

5.17 EDM path without overlap (a) and with overlap (b)[71]

5.18 Array of 86 x 86 quadratic holes[72] (a) and an array of holes with an hon- eycomb structure[73] (b)

5.19 Internal gear[74] (a) and coining die[75] (b) produced by die-sinking

5.20 Improvement method for through hole machining with internal flushing[76].

5.21 Conical orifice created by micro wire EDM[74]

5.22 Die produced by wire EDM[75]

5.23 Production of a micropipe by micro WEDG[68]

5.24 Shape (a) and manufacturing process (b) of a simple shaped microdrill[40]

5.25 Groove[72] (a) and die[74] (b), machined by electro-discharge milling

5.26 Microcavity with 1/8 ball, produced with the uniform wear method[71]. . .

6.1 Relationship between industrial request and state of technology[12]

A.1 Surface roughness measurement conventions

Chapter Introduction

1.1 Background and Motivation

There has been a significant increase in the importance of miniature parts in recent years. The forerunner of this technology was mostly the electronics industry with their need of manufacturing processes for electronic components, like printed circuit boards and inte- grated circuits. The market of microsystem technologies is in general a very fast growing market. According to a study of the European NEXUS organization (Network of Excel- lence in Multifunctional Microsystems), the worldwide market for microsystem technologies is growing at an average rate of 18% a year to a total of $38 billion in 2002[75]. However, the focus of the development is distributed different in certain countries. While the US has for example a focus on parts for micro-electro-mechanical systems (MEMS), equipment for information technology, biomedicine and genetic engineering, Germany dominates in sensor technology for the automotive industry. Japan has traditionally a strong position in fine mechanics and precision engineering as well as in equipment for information technology and consumer goods[77] Until recently, the production of miniature components was focused on technologies, traditionally used in the electronics and semiconductor industry, like etching and other photofabrication techniques. Using these technologies extremely small feature sizes can be produced. Optical lithography for example produces features as small as 0.18 µm and X-ray lithography can be used to produce even smaller features[78]. Table 1.1 gives an overview of some of the methods which can be used for the production of miniature parts. An introduction to these techniques is given in some papers which briefly summarize dif- ferent micromachining methods. A very good paper was published by Masuzawa[12]. The most complete description of different processes is included in the book ”Fundamentals of microfabrication: the science of miniaturization” by Marc J. Madou[13]. Some other papers summarizing different micromachining methods are for example [69, 79-81].

Abbildung in dieser Leseprobe nicht enthalten

Table 1.1: Methods used for micromachining[1].

Because of the steadily increasing demand for 3-dimensional microparts, for example for the development of micromachines, techniques which are capable of creating these features have a growing importance. However, some of the micromachining technologies mentioned in Table 1.1 lack the capability to produce complex 3-dimensional parts with high aspect ratios. The processes, which are best suited to create these parts are for example laser mi- cromachining, mechanical micromachining, microforming, electrochemical machining, and micro electro-discharge machining. Besides the creation of truly 3-dimensional miniature parts with a high aspect ratio, the achievable accuracy, respectively the relative accuracy is always a challenge in micromachining. While the absolute accuracy which can be reached in micromachining is excellent, the relative accuracy is rather bad. In traditional manu- facturing of precision parts relative accuracies of 10−[6] can be attained, but these relative accuracies cannot be reached in micromachining[13]. To clarify about which range of sizes, accuracies, and relative accuracies one talks in micromachining, some examples are shown in Figure 1.1. Micromachining means usually that the size of the parts is smaller than 1 mm, but even feature sizes of less than 1 µm are created. As it is shown in Figure 1.1, 1 µm is the hardly imaginable size of a bacterium. With some processes, surface finishes as good as 1 nm can be created, which is approximately ten times the size of an atom. If the relative accuracy of a house, which is not considered to be a precision part, is compared with the relative accuracy of a lithography based micromachine, it can be noticed that these are approximately the same. The optimum in relative accuracies is usually achieved at sizes, that are bigger than micromachined features. In general, it can be seen that the smaller the micromachined part is, the more difficult is it to achieve a good relative accuracy. Neverthe- less, the absolute accuracy of micromachined features increases with a decreasing size of the machined feature. Other challenges which need to be considered in micromachining are the assembly and handling of the small parts and the stringent requirements for the alignment and positioning accuracy of the tools. However, many of the micromachining techniques mentioned cannot be accounted for specific industrial applications, because they are still in the research and not the industrial stage.

Even though these micromachining processes currently draw a lot of attention in research, there was only little effort put into summarizing and synthesizing the results of this research. Most of the information exists in form of very specific publications, examining only one factor of a certain process. This is unsatisfying if one wants to get information on the capabilities and state-of-the-art of a process or if several processes should be compared. Especially, as these processes are becoming more important in industrial applications, it is crucial for companies to have the possibility to judge the different manufacturing possibilities.

Abbildung in dieser Leseprobe nicht enthalten

Figure 1.1: Size range, absolute accuracy, and relative accuracy in precision machining[13].

1.2 Research Objective, Scope, and Research Plan

The objective of this thesis is to discuss and compare various micromachining processes responsible for creating 3-dimensional miniature parts. Specifically, in this thesis laser mi- cromachining, mechanical micromachining, microforming, and micro electro-discharge mi- cromachining will be considered. These processes were identified to be some of the most important methods for the production of truly 3-dimensional structures with high aspect ratios at the micro-scale. The processes are mainly examined in regards of their process characteristics, capabilities, readiness, and costs. These processes are not the only pos- sibilities to create 3-dimensional miniature parts. Other possibilities like electrochemical machining, ultrasonic machining, ion beam machining, stereo lithography, or electron beam machining also exist, but they are not discussed in this thesis.

To meet the research objective, the following specific tasks are planned. For all the pro- cesses mentioned before, the physical background of the material removal/forming process and the models describing these processes are briefly covered. Specific factors that influence each micromachining process and the differences between the application of each of the pro- cesses at the macro-scale and at the micro-scale are also discussed. This thesis gives the state-of-the-art of the processes in the research area as well as in industrial use. Crucial factors to evaluate the capabilities of each process, like the versatility of shapes that can be produced, minimum feature sizes, machinable materials, possible aspect ratios, achievable surface finish, machining strategies, and the machining speed are discussed. This is mostly done using examples from former research projects. Commercially available machines and tools are covered and, because of the importance for the industrial use of these micromachin- ing processes, the costs for applying each method in an industrial application are discussed. The general objective is to collect and structure all the information mentioned above to draw a picture of each process, which is as complete as possible, and give an extensive overview of existing publications, covering these processes.

1.3 Thesis Content

This thesis is structured in an introduction and 5 more chapters. The basic content of these chapters is summarized below.

Chapter 2 discusses laser micromachining. After identifying CO2 lasers, Nd:YAG lasers, Excimer lasers, and Femtosecond lasers as the most important laser for micromachining, these lasers are examined in more detail regarding the factors of interest mentioned above.

Furthermore, the laser ablation process is examined and the factors influencing the machining process and characterizing the different laser types - wavelength, laser spot size, average laser beam intensity, depth of focus, pulse duration - are discussed.

Chapter 3 focusses on the research covering mechanical micromachining processes, includ- ing microdrilling, microturning, and micromilling. For each of the methods the influence of important factors like for example the feedrate is discussed. Differences between the conven- tional process at the macro-scale and micromachining are examined and process models are introduced.

In Chapter 4 microforming processes are briefly examined. After introducing the reader to the different microforming techniques - extrusion, cold forging, deep drawing, embossing/coining, air bending, and blanking/punching - factors distinguishing microforming from conventional forming are discussed.

Chapter 5 covers the micro electro-discharge machining processes. The versatile tech- niques which can be applied at the micro-scale, like electro-discharge die sinking, electro- discharge drilling, wire electro-discharge machining, wire electro-discharge grinding, electro- discharge grinding, electro-discharge milling, and the uniform wear method are discussed. Furthermore, the principle of the material removal process in electro-discharge machining is explained.

Chapter 6 summarizes the information obtained in the previous chapters. An attempt to compare the capabilities of the different methods is made and important issues in research and industrial applications are stressed.

Appendix A introduces the different surface roughness measures used in this thesis.

Finally, Appendix B gives an overview of some specifications of industrial micromachining equipment discussed in this thesis.

Chapter 2 Laser Machining

2.1 Introduction

Laser Micromachining is a technology that was used first in the 1960s in the cutting of trim grooves on conventional resistors and drilling small holes in diamond[14]. Since that time, laser technology has developed in an impressing way. New kind of lasers and new micromachining techniques have been developed.

Currently CO2, solid-state lasers (e.g. Nd:YAG), copper vapour lasers, diode lasers, ex- cimer lasers and femtosecond (ultrashort) lasers are used for micromachining[2]. In this chapter the most important ones for micromachining (CO2, Nd:YAG, excimer and femtosec- ond laser) are examined in more detail. An overview of typical peak power, pulse length, and wavelength ranges for Nd:YAG, excimer, and femtosecond lasers (titanium:sapphire) is given in Figure 2.1.

The micromachining techniques used are either direct writing (direct focusing) or mask projection. In direct writing an interface to CAD-data is possible, allowing a fast machining of complex structures and a rapid prototyping process. Whether direct writing or mask pro- jection is used depends mainly on the laser used and the shape one wants to manufacture. In

Abbildung in dieser Leseprobe nicht enthalten

Figure 2.1: Overview of typical laser characteristics for Nd:YAG, excimer, and titanium:sapphire lasers[14].

general, mask projection is mainly used in combination with excimer lasers. The advantages and disadvantages of these techniques are summarized in Table 2.1.

Abbildung in dieser Leseprobe nicht enthalten

Table 2.1: Comparison of direct writing method with mask projection method[2].

A laser system for material processing typically consists of the beam source, a focusing 8 system, a motion control for the workpiece, the computer control and a video surveillance system. In this chapter the advantages, disadvantages and capabilities of the different lasers and techniques are pointed out. Because of the extent of this thesis it is not possible to explain the functioning and the material removal process for all the different lasers in detail. Only the factors which are important for the machining process itself are discussed in this paper.

In a very simplified way, the material removal process can be explained as follows[19]:

- Heating of the material by absorption of the laser energy
-Melting of the material
-Vaporization of the material

This process is illustrated in Figure 2.2. The absorbtion of the laser energy occurs first at the electrons, which transfer the absorbed energy to the lattice of the material. In consequence, the lattice starts oscillating and heat conduction is taking place. The material starts melting and is still absorbing energy from the laser beam. Finally the evaporation temperature is reached and the material is removed from the work spot as a vapor by the expulsion of melt in an explosivelike boiling [14, 19]. Because usually some of the material re-solidifies, there is some rim of material and debris observable when machining metals[14]. Furthermore, most of the time a heat-affected zone is created at the work spot.

However, not all of the descriptions above are correct for ultrashort lasers or lasers with an ultraviolet wavelength. The ablation process of these lasers is different and described in more detail in chapter 2.6.2. This is done because of its importance to the specific characteristics of femtosecond laser machining.

A drawback of all the laser micromachining techniques is that it is not possible to obtain completely vertical walls. This is especially true for high aspect ratio features. The larger the aspect ratio is, the larger is the tapered angle[5]. A typical tapered angle for excimer

Abbildung in dieser Leseprobe nicht enthalten

Figure 2.2: Laser beam interaction and material removal.

laser micromachining is 7o, but it can be increased by using low fluence or reduced to approximately 2o by using a high fluence. The tapered angle can be even more reduced by simple double sided drilling of a hole. This results in an ”hourglass” shaped feature[3].

2.2 Laser Parameters

There are several laser parameters existing that can be controlled and have an influence on the micromachining process and the machining quality. The most important ones for micromachining are the wavelength, spot size, average beam intensity, depth of focus and laser pulse length[13]. These different parameters are the reason why it is difficult to exactly determine the surface finish and tolerances that can be obtained with laser micromachining. These values are very application dependent. In this section the five parameters and their general influence on micromachining operations will be discussed.

Laser Wavelength

The wavelength is an important parameter, because the material removal process is different for various wavelengths. At very short wavelengths (~ 200 nm) direct photochemical bond breaking plays an important role in the ablation process. Photons emitted at this wavelength have often a high enough energy that they might cause a direct bond cleavage in the solid and, therefore, initiate a rapid decomposition of the material into highly volatile molecular fragments. This process is not possible for longer wavelengths (~ 300 nm). In this case the ablation is a thermal process[13].

Laser Spot Size

One of the main objectives of micromachining is to create very small features. Therefore it is necessary to focus the laser beam on a very small spot size. Using a lens with no spherical aberration (”perfect lens”), the minimal spot diameter that can be obtained is determined by the diffraction limits of the imaging system. The correlations between the influencing factors are shown in equation 2.1 [13, 14]:

where dmin = minimum diameter of the focused beam

λ = the wavelength of the radiation

M[2] = the beam mode parameter (it measures how close

the actual laser beam is to the theoretical beam

profile, with M[2] = 1 signifying a perfect match

d0 (mm) = the diameter of the beam at the focusing lens

f = the focal length of the lens

Therefore in practice a small spot size is obtained by using a short wavelength.

Average Laser Beam Intensity

The laser intensity (W/cm[2] ) is the peak power divided by the focal spot area, and the peak power is the pulse energy divided by the pulse duration[13]. The intensity is crucial for the time it takes the beam to melt the material. With I = 10[9] W/cm[2] on steel, for example, 300 ns are needed to reach the melting point[14].

The average on-axis laser beam intensity is[13]:

where I = the average on-axis laser beam intensity

P = the total laser power in watts

W[2] [0] = the Gaussian beam radius - the radius at which

the intensity has decreased to 1/e[2] or 0.135 of its value on the axis

Depth of Focus

The depth of focus is defined as the distance between the values where the beam is 2 times larger than it is at the beam waist. The depth of focus can be computed using equation 2.3[13].

Abbildung in dieser Leseprobe nicht enthalten

In order to process material using a very short depth of focus, the surface topology needs to be very flat Laser Pulse Duration The fifth important parameter for laser micromachining is the pulse duration used. For long pulses (> 10 ps) the heat deposited in the workpiece diffuses away during the pulse duration. This is correlated to numerous undesirable side effects, including reduced efficiency, surface debris and a heat-affected zone. Short wavelengths are a possibility to reduce or even eliminate these effects. This is the property femtosecond lasers exploit and it will be discussed in more detail in Chapter 2.6.

2.3 CO2 Lasers

2.3.1 Introduction

Because of their high status of development, their reliability, their rather low prices and high efficiency, CO2 lasers are one of the most common lasers in industry. They operate in the infrared part of the electromagnetic spectrum, at a typical wavelength of 10.6 µm[3], and a typical pulse duration of 200 µs[14]. CO2 lasers are available with maximum power outputs of up to 500 W and pulse repetition rates up to 20 kHz[82].

2.3.2 Capabilities

CO2 lasers are operated in a direct focusing technique for drilling or cutting of the target material. While a theoretical feature resolution of 10 µm can be obtained, it is in practice approximately 50 to 75 µm. The maximum attainable aspect ratio is about 100:1[3].

The material removal is performed by overloading the target surface thermally. The high penetration depths of most 10.6 µm absorbing materials results in a fast material removal. However, CO2 lasers do not produce high enough peak power densities to efficiently machine a lot of metals, especially highly reflective metals as gold, aluminium, or copper[82]. For some materials, only around 20% of the laser energy is absorbed[14]. Therefore, CO2 lasers are definitely more appropriate for machining materials with a transmission plateau in and around visible frequencies. These materials are glasses, semiconductors, quartz, and oxide- based ceramics[83]. CO2 lasers are nevertheless not very suitable for the machining of diamond, which has a transmission plateau in the visible area as well, because the diamond is heated too much and turns to graphite during the machining process[82].

Another drawback of CO2 lasers is the formation of a large heat-affected zone. A flow of the surrounding material and crack formation due to the thermal shock result from the high temperature reached during the machining process[84]. In general the overall machining quality of CO2 lasers is much lower than the quality of excimer or femtosecond lasers. However, there have been a number of advances in recent years. For example, a q- switched CO2 laser has been developed, resulting in pulse durations shorter than 100 ns[85]. This results in a significant reduction of the heat-affected zone when machining plastics. Therefore finer cuts and holes can be produced. The q-switch furthermore produces higher peak powers and extents the range of machinable materials to fiberglass composites and ceramics. Another advantage of q-switched CO2 lasers is that they support repetition rates up to 100 kHz[85].

For the processing of Kapton, Collins confirmed with her experiments that the pulse width dramatically affects the diameter of laser drilled holes. Using a constant repetition rate and the same number of pulses, the diameter of a drilled hole increased nearly linearly when the pulse width was increased from 15 µs to 1500 µs[86].

2.3.3 Industrial Applications

The most important commercial application of CO2 lasers in micromachining is currently the drilling of microvias for printed circuit boards (PCBs). Because of its high dielectric material removal rate, it is the most economical solution for large-scale production of mi- crovias of 150 µm and below [82,87]. Drilling rates of up to 20,000 surface blind via holes per minute on thermally curable, resin screen coats, have been reported[82]. However, because PCBs consist of a copper layer and the dielectric material, very often a dual strategy is used, in which a ultraviolet laser is taken to simultaneously machine the copper layer more efficiently[87].

2.3.4 Costs

The capital investment one needs to make for a CO2 micromachining laser system is approx- imately $150,000 to $250.000[88]. CO2 lasers are, regarding the investment as well as the operating and maintenance costs, the cheapest of the four different laser systems compared in this thesis.

2.4 ND:YAG Lasers

2.4.1 Introduction

Another popular laser for micromachining, which has a high status of development, is the solid state Nd:YAG laser. This laser operates at a fundamental wavelength of 1064 nm, but non-linear crystals enable it to be frequency converted to 532 nm, 355 nm, or 266 nm. These four states are called first, second, third and fourth harmonic. The conversion efficiency is decreasing for the conversion to smaller wavelengths [3, 89]. Today, mainly diode-pumped solid-state lasers (DPSS), where the gain medium is pumped by a laser diode, have big success. These lasers have, compared to conventional flashlamp-pumped solid-state systems, the advantage that they are up to 5 times more efficient. Other advantages are their better beam quality and the reduced operating costs because of the long lifetime of the diodes (~10,000 hours) [89-91]. DPSS lasers are Q-switched and have pulse lengths typically in the order of 10 - 50 ns. Their focusable spot sizes range in the microns, and they produce intensities on the surface of the material of the order of 10[10] - 10[12] W/cm[2] [89]. The repetition rates of commercial available Nd:YAG lasers, for example, the Lambda Physik Gator series, is approximately 10 kHz and the amplifier output is, depending on the wavelength, up to 30 W[16].

To optimize the settings of the different laser parameters (peak power, average power, pulse duration, pulse frequency, cutting speed) for a certain application, Chen et al. gave some guidelines and a proposed a theoretical model to determine these settings[92].

2.4.2 Capabilities

The removal process of Nd:YAG lasers depends primarily on the material, the pulse duration, the wavelength, and the laser fluence. The removal of metals is basically a thermal process via fast melting and evaporation. In order to maintain a high accuracy and a small heat affected zone, the laser fluence for the machining of metals should be limited to a few 10 J/cm[2]. At this point material removal rates of about 1 µm per pulse for stainless steel or titanium, and 10 µm per pulse for aluminium are obtained [16, 93].

For laser it is normally the case that the energy deposited into the material needs to exceed a certain threshold in order to remove material. This threshold is called the fluence threshold Fth [J/cm[2] ]. The fluence threshold for alumina is at around 2 J/cm[2] for 355 nm, and for PTFE (e.g. Teflon) it is one order of magnitude higher at 20 J/cm[2]. The removal rates are at 1 µm per pulse for alumina and 10 µm per pulse for PTFE. PTFE cannot be machined with 1064 nm[93].

As it is shown in Figure 2.3 a), for surface micromachining, material removal rates only increase slightly as the laser fluence is increased more than one order of magnitude above these thresholds[16]. Increasing the laser fluence beyond this threshold increases the plasma density due to photoionization and inverse bremsstrahlung (Joule heating process), and does not result in a faster material removal. But as shown in Figure 2.3 b) and c), this property changes for an increase in the thickness of the sample. For a thickness of 0.870 millimeters, the necessary laser fluence to penetrate the entire sample of stainless steel is approximately 100 J/cm[2]. At lower laser fluences, the material removal terminates at a certain depth due to attenuation of the beam in the column of vapor and plasma that fill the drilling hole[16].

Another effect that appears for high aspect ratio drilling is shown in Figure 2.3 c) and Figure 2.4. As the sample thickness approaches 1 mm, the removal rate becomes more and more wavelength dependent. The difference in the removal rate for a one mm thick stainless steel sample is nearly on order of magnitude. Even though the removal rate for surface micromachining is nearly independent of the wavelength, the attenuation of the beam in high aspect ratio holes is much higher for longer wavelengths [15, 16].

As already mentioned before, short pulse durations are usually used for micromachining. Short pulses result in a smaller plasma plume during the pulse, and therefore in less plasma

Abbildung in dieser Leseprobe nicht enthalten

Figure 2.5: Applications of Nd:YAG lasers, depending on laser parameters[15].

Nd:YAG lasers can be used to machine a lot of materials including metals, ceramics, diamond, sapphire, and silicon. Under certain conditions, aspect ratios up to 100:1 can be achieved[3]. The strength of these lasers is definitely the machining of hard metals. Using a commercial laser of the Gator series, holes with an aspect ratio of 50 have been drilled into 1.46 mm-thick stainless steel (diameter 31 µm). The resulting entrance and exit hole is shown in Figure 2.6 (different scales!). The drilling time for this hole was 0.4 seconds[17].

Hartke et al. report that they machined a 300 µm deep and 15 µm wide (aspect ratio: 20) groove in aluminium, using a Nd:YAG laser at 532 nm wavelength[94]. Zhang et al. used a Nd:YAG laser with a wavelength of 355 nm, a pulse duration of 50 ns and a repetition rate of 2 kHz to perform some experiments in the laser drilling of holes

Abbildung in dieser Leseprobe nicht enthalten

Figure 2.6: Entrance side (31 µm) and exit side (26 µm) of a percussion-drilled hole in 1.46 mm stainless steel (10W, 10kHz, 355nm)[17].

in copper[95]. The minimum beam size of the used laser was found to be 4.558 µm and the laser beam intensity had a Gaussian distribution. Among other experiments, Zhang et al. drilled holes at intensity levels of 6 - 10[8] W/cm[2] and 2.5 - 10[9] W/cm[2]. At 6 - 10[8] W/cm[2] the heat affected zone was approximately 1 µm and the diameter 3.3 µm. The hole drilled at an intensity level of 2.5-10[9] W/cm[2] had a heat affected zone of about 3 µm and its diameter was 7.3 µm. The surface quality was slightly better using the lower intensity. The experiments were performed using only 10 respectively 50 laser pulses. The depth of the holes is higher as the laser intensity used is higher[95].

In another experiment, Heyl et al.[18] developed a laser ablation machine named LAM 66, with a Nd:YAG laser operated at 355 nm functioning as beam source. The LAM 66 is capable of manufacturing microstructures with a minimum feature size of 10 µm, and an accuracy of 1 µm in a maximum working area of 100 x 100 x 5 mm[3] (length/width/height).

The machine can produce features with 2[1] dimensions, undercutting of structures is not possible. The principle setup of this machine and a structure machined in WC/Co (tungsten carbide sintered in a cobalt matrix), which was produced by computer-driven laser ablation is shown in Figure 2.7. A data interface allows to import computer designed 3D models into the LAM 66. With optimized settings an average surface roughness Ra of 0.16 µm and maximum surface deviation Rz of 0.7 µm can be obtained. In general, better surface qualities are reached by using low laser fluences[18].

Abbildung in dieser Leseprobe nicht enthalten

Figure 2.7: Setup of LAM66 and an example structure, machined in Wc/Co with a depth of 150 µm[18].

2.4.3 Industrial applications

An industrial application for Nd:YAG lasers is as for CO2 lasers, the drilling of vias in printed circuit boards. The advantage of solid state lasers over CO2 lasers is their capability to penetrate the dielectric material and the copper layer, their smaller focused spot size, and their reduced thermal damage. Vias with a diameter down to 25 µm can be drilled[96].

In general every process, where small diameter holes or features need to be machined in hard material, is a good application for Nd:YAG lasers.

2.4.4 Costs

The capital costs for Nd:YAG laser systems range between $150,000 and $300,000[88]. Their operating and maintenance costs can be approximated to be $4 to $5 per hour[97].

2.5 Excimer Lasers

2.5.1 Introduction

The third popular laser for industrial micromachining applications is the excimer laser. The big advantage of excimer lasers is that they emit light in the ultraviolet spectrum (157 - 351 nm). This results in a small penetration depth of the beam, and therefore an energy absorption near to the surface [5, 19]. The short wavelengths of excimer lasers allow a high resolution.

Excimer lasers are gas lasers that are operated using one of six noble gases, the characteristics of which are shown in Table 2.2. The beam profile of excimer lasers is, in contrast to the other kind of lasers, not Gaussian. It is only Gaussian on the short axis and flat-topped along the long axis (usually about 10 mm to 25 mm)[20]. However, as the gas fill ages, the beam properties change. The beam grows along the short axis, resulting in a flat-topped profile for both axis[3]. Because of this poor spatial coherence, direct focusing is rather unattractive. Instead a mask projection technique is used[20].

Abbildung in dieser Leseprobe nicht enthalten

Table 2.2: Excimer gas mixtures[3].

The pulse duration of excimer lasers depends marginally on the gas filling used. Pulse durations are typically in the order of 20 ns. Repetition rates for commercial lasers are up to 2 kHz, but range often around 200 Hz. The attainable pulse energy for systems with very

high repetition rates (~ 1 kHz), is of the order of one magnitude lower than the energy for low repetition rates (~ 10 Hz)[19].

Furthermore, excimer lasers can be used for surface treatment, but this is not discussed in this thesis. Introductional information on this use is provided by Boulmer-Leborgne et al.[98].

2.5.2 Techniques

There are different mask projection techniques to machine micro-features using an excimer laser. For all the techniques, the beam is usually homogenized to make a ”flat-top” beam. This is done because the ablated depth depends on the energy density of the beam at a certain point[20]. The homogenized beam is imaged onto the sample in a certain shape, which is specified by a mask usually made from chrome-on-quartz or a thin metal sheet. The features of the mask do not need to be as small as the structures to be produced because the projection lens usually de-magnifies the mask pattern on the workpiece by the factor x4, x10 or x30. So the mask need not have the high resolution features, and the danger of damaging the mask with the laser beam is lower, too. This is due to the reduced energy density because of the de-magnification[20]. Depending on the material and the laser energy, a typical field size at the workpiece is approximately 5 x 5 mm[2] [99].

Static mask and workpiece Static mask projection is used to produce discrete features when the pattern is small and simple, or consists of regular, repeating patterns. This technique is mostly applied in the production of holes (see Figure 2.8). Two extensions of static mask projection are the step- and-repeat processing and the indexed mask projection. Step-and-repeat processing means that static mask is used to produce a structure and then the laser is turned off, the sample is moved laterally in X or Y direction, and then the laser is fired again. This technique produces a simple, repeating structure. Indexed mask projection is used to produce structures that are slightly more complicated. After creating a feature with a static mask, the laser is again turned off, but now another mask pattern is moved under the laser beam. Usually a stepped structure is created [19, 20].

Figure 2.8: Static mask projection at the example of a hole mask (left) for the simultaneous production of a hole array (right)[19]

Moving mask

To produce structures with a varying depth profile, the moving mask technique is used. The aperture moves across the laser beam in an exactly defined matter. This results in a continuous change in the amount of energy to which the workpiece is exposed, and therefore in a depth gradient in the sample. If the settings of mask aperture, mask motion, workpiece positioning and pulse sequence are chosen correctly, complicated features as shown in Figure 2.9 b) and c) can be produced[20].

Moving workpiece

Another mask projection technique is the movement of only the workpiece. This technique is commonly used to produce micro-channels or micro-grooves. The shape of the groove is

Abbildung in dieser Leseprobe nicht enthalten

Figure 2.9: Linear ramps produced in PET at 248 nm, using the moving mask technique[20]

defined by the shape of the mask (e.g. triangle, ’T’-shape, circle), the movement speed of the workpiece and the other typical characteristics such as the fluence. The mask and the groove created using this triangular mask, are shown in Figure 2.10 a). The technique of a moving workpiece is used as well to create multi-shaped 3D micro-structures, for example pyramids (see Figure 2.10 b)), cones, or lenses. This is done by moving the workpiece in two dimensions [19, 20].

Abbildung in dieser Leseprobe nicht enthalten

Figure 2.10: Moving workpiece: a) General principle[19] ; b) Pyramids machined using the moving workpiece technique[21]

Moving mask and workpiece

The combination of a moving mask and a moving workpiece, also called synchronized scanning, is used where the pattern to be produced is large, non-repeating and cannot be produced by any of the three techniques explained before. The drawback of this technique is that the whole pattern to be produced has to be contained in the mask[20].

Direct writing using mask projection

The advantages of a direct writing system, for example the possibility of interfacing the laser system directly to a CAD system, can be imitated by using a circular, stationary mask. The circular mask defines a beam spot of a certain diameter and therefore the system can be used as a direct writing system that can be directly interfaced to the CAD generation of the pattern.

Hole Area Modulation (HAM)

A new excimer laser configuration used for the creation of three dimensional features is Hole Area Modulation. The information about the pattern that should be produced is included in the mask for this technique. The mask has holes with different diameters and the workpiece is usually moved randomly within an area corresponding to the pitch of the array. The work- piece is moved randomly to ensure an equal energy distribution over the whole workpiece, but sometimes computed movement paths are applied, too. The relative machining depth is defined by the different diameters of the holes in the mask. The bigger the diameters are, the more energy reaches the workpiece during the machining (Figure 2.11). HAM is capable of machining three dimensional features very fast and with a lower energy loss than using traditional masking techniques. The drawbacks of HAM are that the achievable resolution is limited by the mesh size, and that, because of the random movement of the workpiece, no sharp corners can be machined[22].

Abbildung in dieser Leseprobe nicht enthalten

Figure 2.11: HAM - Expected section of the cavity machined with a mask with holes of two different diameters[22]

Micro Lens Array (MLA)

A system that does not use a mask and therefore does not have the drawback that most of the laser energy is lost at the pattern mask, is the micro lens array. Sekizawa et al. demonstrated a system where they arranged 2,000 micro lenses to focus the inconstant laser beam to spots. Doing this they have been able to simultaneously create 2,000 holes with a diameter of 50 µm in a resin layer[100].

2.5.3 Capabilities

Excimer lasers are the lasers creating the smallest features and the best tolerances that are already in commercial use. Even though the technique is not completely matured yet, there have been huge developments in recent years. Excimer lasers become increasingly visible in industrial settings (24 hours/day; 7 days/week operation)[3].

Excimer lasers are capable of machining most of the materials, including polymers, ce- ramics, glass, metal, sapphire and diamond [5, 14]. Using F2 lasers with a wavelength of 157 nm even difficult to machine materials as PTFE (e.g. Teflon) and quartz glass have been machined with good results. Figure 2.12 shows the difference in the machining quality when wavelengths of 193 nm and 157 nm are used, respectively . For fused silica and PTFE, machining at 193 nm still produces a lot of thermal damage and debris, while the results using 157 nm are much better [23, 101].

Abbildung in dieser Leseprobe nicht enthalten

Figure 2.12: PTFE, processed at 193 nm (a) and 157 nm (b). Fused silica processed at 193 nm (c) and 157 nm (d).[23]

Since the applied photon energy of excimer lasers (Table 2.3) is similar to the level of molecular bonds in plastics, these and similar materials are the ideal targets for excimer laser micromachining. The ablation process is often related to as a non-thermal process, because the removal of the material is caused by direct bond breaking and not by heating and vaporization of the material [4, 102]. Therefore hardly any heat-affected zone is observ- able. This results in a high dimensional accuracy and less defects in the surface layer[12]. Even though metal is not the most appropriate material for using excimer lasers, it can be

processed with a higher power, too. But in this case the beam-matter interaction involves some thermal phenomena[103].

Abbildung in dieser Leseprobe nicht enthalten

Table 2.3: Energy per photon at different excimer wavelengths[4].

Some processing characteristics of micromachining using an excimer laser are shown in

Table 2.4. Typical ablation rates for an excimer laser are approximately 0.1 µm/pulse[104].

Abbildung in dieser Leseprobe nicht enthalten

Table 2.4: Ablation rate using a KrF excimer laser[5].

For metals and material that behave to some extent similar to metals (e.g. Silicon), some more effects determine the drilling process. As shown in Figure 2.13, there is hardly any effect of increasing the fluence over the threshold fluence, but a reduction of the beam diameter increases the removal rate dramatically. This effect especially applies when the beam diameter is reduced below 20 µm [24, 105]. The reason for this is that a large part of the drilling processes for these materials is done by melt ejection. As it is explained in more detail by Asada et al.[24], melt ejection is much larger for smaller diameters.

One of the most important figures for precision micromachining is the resolution that is achievable with a certain machining tool. Excimer lasers have a theoretical resolution ranging between 0.2 and 0.4 µm. However, the practical imaging resolution is limited to

Abbildung in dieser Leseprobe nicht enthalten

Figure 2.13: Relationship between drilling rate of silicon and laser fluence at different beam diameters. Dots are experimental values and lines are calculated[24]

approximately 1 - 5 µm [3, 104]. The attainable aspect ratio is up to more than 100:1. In contrast to CO2 and Nd:YAG lasers, there is primarily only a small recast layer around the edge. Burring and thermal effects are hardly observed[3]. The hole shown in Figure 2.14 was drilled in a 0.5 mm thick alumina ceramic plate. This was done by mask-illumination technique at a wavelength of 248 nm and de-magnification factor of 15. It took 22.4 seconds to produce the hole[25].

A good illustration of the advantages of laser micromachining over mechanical micromachining is shown in Figure 2.15. This figure shows a 100 µm diameter hole, drilled in 75 µm thick high-density polyethylene, using a mechanical twist drill bit (a) and a KrF excimer laser (b)[21]. The better quality of the laser drilled hole is obvious.

[...]