- Email: [email protected]
CO2 laser micromachined crackless through holes of Pyrex 7740 glass
International Journal of Machine Tools & Manufacture 50 (2010) 961–968
Contents lists available at ScienceDirect
International Journal of Machine Tools & Manufacture journal homepage: www.elsevier.com/locate/ijmactool
CO2 laser micromachined crackless through holes of Pyrex 7740 glass C.K. Chung n, S.L. Lin Department of Mechanical Engineering, and Center for Micro/Nano Science and Technology, National Cheng Kung University, Tainan, Taiwan
a r t i c l e in fo
abstract
Article history: Received 19 February 2010 Received in revised form 8 August 2010 Accepted 9 August 2010 Available online 13 August 2010
The drilling of glass through holes with a high aspect ratio is crucial for microsystems application, especially in the inlet/outlet connection of microfluidic devices for biological analysis or for the anodic bonded silicon-glass ones. Traditional glass drilling using mechanical processing and laser processing in air would produce many kinds of defects such as bulges, debris, cracks and scorch. In this paper, we have applied the method of liquid-assisted laser processing (LALP) to reduce the temperature gradient, bulges and heat affected zone (HAZ) region for achieving crack-free glass machined holes. The nominal diameters of circles from 100 to 200 mm were drawn for through glass machining test. Through-hole glass etching can be obtained by LALP for 10 passes of circular scanning in several seconds on conditions of a 6 W laser power, 76 mm spot size and 11.4 mm/s scanning speed. The ANSYS software was also used to analyze the temperature distribution and thermal stress field in air and water ambient during glass hole machining. The higher temperature gradient in air induced higher stress for crack formation while the smaller temperature gradient in water had less HAZ and eliminated the crack during processing. CO2 laser micromachining under water has merits of high etching rate, easy fabrication and low cost together with much improved surface quality compared to that in air. & 2010 Elsevier Ltd. All rights reserved.
Keywords: Pyrex glass CO2 laser Water-assisted Etching Through hole
1. Introduction Glass materials have widely been used in the application of optical and optoelectric devices and microfluidic and bio MEMS. Glass micromachining, especially deep etching of Pyrex glass, was studied by many researchers using wet hydrofluoric acid (HF) based solution [1–3], dry plasma etching [4] or electrochemical discharge microdrilling [5]. HF based solution with the assistance of photolithographic masking can selectively etch glass up to 500 mm deep but it encounters problems of large lateral undercut and very low etching rate of about 12.9 mm/h [1]. The masking materials with the problem of pinhole formation often existed during deep glass etching [2]. The annealed PECVD amorphous Si among different masking materials of Cr/Au, Cr/Cu and itself was tested to be the best one for glass etching up to 200 mm in HF (49%) solution [3]. Deep etching of Pyrex glass using SF6 plasma in a high-cost ICP system also exhibited a low etching rate of about 0.6 mm/min [4]. Electrochemical discharge microdrilling had one drawback: heat-affected zone (HAZ) being left on the microdrilled hole surface [5]. Recently, laser has widely been studied in micromachining. Ultra-short laser systems, picosecond to femtosecond lasers, were used for the fabrication of drilled holes and shaped microstructure in metal materials [6–8] or silicon wafers [9–11]. The ultra-short laser has the advantage of a small heat
n
Corresponding author. Tel.: 886 6 2757575x62111; fax: 886 6 2352973. E-mail address: [email protected] (C.K. Chung).
0890-6955/$ - see front matter & 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.ijmachtools.2010.08.002
affected zone (HAZ) on the surrounding material due to the minimized heat conduction during a short pulse. Several publications used the assisted method to etch transparent silica glass at low repetition rate or higher output laser energy [12–14]. But the expensive femtosecond laser or excimer laser [15] used for etching glass had low etching rates together with the high cost, especially for deep through-wafer etching. The traditional defects of bulge, debris, cracks and scorch often occurred during laser processing to influence the fabrication quality for application. Cheng et al. [16] reported acid solution assisted UV-laser wet etching glass substrate for microfluidic application but the etching rate was still low, together with the occurrence of traditional high-bulge defects. A liquid-assisted laser processing (LALP) has been applied to the welding, cleaning or cutting of metals, alloys and crystalline ceramics [17,18]. The drilling of through holes with a high aspect ratio is important for potential industrial application. Although CO2 laser was widely used for transparent material ablation [19], it was rarely used to directly etch through amorphous Pyrex glass at room temperature in air due to the thermal stress induced crack problem [20]. Ogura and Yoshida [21] reported crystalline and amorphous glass hole drilling using CO2 laser and found that crystalline synthetic quartz had the best drilling quality together with through-hole drilling while amorphous Pyrex glass had the worst result together with the highest bulge around the hole as well as no through-hole cross section being demonstrated. Yen et al. [22] reported the substrate heating method to improve the defects of CO2 laser ablation on glass for microfluidic application. But no
962
C.K. Chung, S.L. Lin / International Journal of Machine Tools & Manufacture 50 (2010) 961–968
through-hole etching was shown and such method would affect the cutting performance and increase instrument complexity. Chung et al. [23] reported the linearly through-wafer etching of Pyrex glass using liquid-assisted CO2 laser process but with no through-hole etching. Barnes et al. [24] proposed the water assisted CO2 laser cutting alumina specimen compared with laser processing in air, then a finite element approach was used to analyze the temperature and thermal stress distributions during both cuttings. Chung et al. [25] used the combination of waterassisted CO2 laser processing for ablating glass slides and the modified thermal bonding for capillary-driven bio-fluidic application. In this paper, we have demonstrated the effective method of through-hole etching of amorphous Pyrex glass using the lowcost CO2 laser based on LALP technology. The liquid form of water was used to obtain good quality of through hole glass etching for the sake of cooling and cleaning. In order to further investigate the temperature distribution, maximum temperature and thermal stress during the drilling machining process, a simple thermal model was adopted for the analysis of thermal behavior during machining. ANSYS software was used to solve this transient problem. The water cooling effect and debris removal mechanism were also discussed for drilling quality improvement.
power of 6 W and scanning speed of 11.4 mm/s at different scanning numbers from 1 to 10 passes were applied to glass etching. Both kinds of water depth were controlled at about 0.5 and 1.0 mm above glass surface by the total volume method whose depth is proportional to the volume at a constant area of holder. In order to keep the glass surface always wet during ablation, 5 ml of hydrophilic surfactant agent of lauramidopropyl betaine (RCONH(CH2)3N + (CH3)2CH2COO–) was added to avoid the water layer broken into drops. Traditional laser processing in air was performed as a reference sample. The optical microscope (OLYMPUS BX51M, Japan) was used to examine the morphology after laser processing. The profile and morphology of lasermachined glass channels were examined by the alpha-step (a-step) profiler (Kosaka Lab, ET3000, Japan). ANSYS software was used to investigate the temperature distribution and thermal stress field in air and water ambient during glass hole machining. The laser beam focusing on the plane surface maintains a constant TEM00 mode, and laser intensity distribution is a Gaussian function model and the heating phenomena due to phase changes are neglected. The beam of CO2 laser is regarded as a surface heating source but the surface of Pyrex 7740 glass without laser heating and the superficial heat irradiation is negligible.
3. Results and discussion 2. Fabrication and experimental procedures Fig. 1 shows the schematic diagram of liquid-assisted CO2 laser processing setup. It includes the CO2 laser source, a reflective mirror, a focal lens and a sample immersed in water. The test sample is amorphous Pyrex 7740 glass with an average thickness of 500 mm. The glass was submerged in water in a container at room temperature, and then laser beam was performed to etch glass at proper parameters. The working liquid selected was water due to its chemical inertness. The commercial available air-cooled CO2 laser equipment (VL-200, Universal Laser system Inc., USA) was used with a maximum laser power of 30 W. The CO2 laser has a wavelength of 10.6 mm and a TEM00 output of beam. Also, the CO2 laser uses a sealed-off, RF excited, slab design and a multipass, free space resonator for a good quality beam with an M2 value of 1.4 70.2. The maximum laser scanning speed was 1140 mm/s and the largest working area was 409 304 mm2. The focal length of the lens is about 38.1 mm and the smallest beam spot size of the commercial product after standard verification could reach 76 mm, which is defined as the double distance from the center of Gaussian intensity distribution to that with 1/e2 maximum intensity. The laser spot was focused on the top surface of Pyrex glass during LALP. The laser power, scanning speed and the number of drilling passes were controlled to get a better processing result. A computer aided design program of CorelDraw software was available to set the experimental parameters for auto-controlled processing. The constant laser
Fig. 1. Schematic diagram of CO2 laser micromachining setup. It includes the CO2 laser source, a reflective mirror, a focal lens and an experimental sample immersed in water.
Fig. 2(a)–(d) shows the top-view and cross-sectional optical micrographs of CO2 laser etched glass holes in air with drawn diameters of 100 (Fig. 2(a)–(b)) and 200 mm (Fig. 2(c)–(d)), respectively, at a power of 6 W, a speed of 11.4 mm/s and 10 passes. Cracks and scorches frequently appeared on the glass surface during laser processing in air. The cracks occur prior to through-hole glass etching because of the high thermal stress [20] while the scorch formation is mainly from the huge amount of accumulated heating energy for burning, especially more in circular etching than in linear cutting [26]. Since the focal length of the lens is about 38.1 mm and the laser beam intensity is a Gaussian distribution, the etch rate is higher at the center than that at the rim of hole. In addition, a clear HAZ area can be found in Fig. 2(b) and (d) due to the existence of temperature gradient. The higher temperature gradient leads to the larger HAZ area even cracks because the drawn diameter of 100 mm has larger thermal accumulation energy than that of 200 mm. The CO2 laser micromachined glass hole using LALP on a drawn diameter of 100 mm is shown in Fig. 3(a) and (b) in terms of top view and cross section. The liquid is water and about 0.5 mm above the surface of glass. The laser operation parameters are the same as those in air in Fig. 2(a) and (b). No cracks and scorches appear. The HAZ area by LALP is smaller than that in air. The central white spot indicates the through-hole etching and it has a smaller diameter than the total circle due to the tapered profile (Fig. 3(b)) formed by laser ablation. Similar results can be obtained by etching of a hole in larger diameter. Fig. 4(a) and (b) shows the top-view and cross-sectional optical micrographs of the CO2 laser etched Pyrex glass with a drawn diameter of 200 mm using LALP, respectively, at a power of 6 W, a speed of 11.4 mm/s and 10 passes. The water depth is the same 0.5 mm above the surface of glass. Also no crack or scorch is formed under LALP. The tapered profile is more diffusive than that in the small diameter hole. The water depth will affect the efficient laser power and etching rate. Fig. 5(a) and (b) shows the etching depths of the holes in drawn diameters of 100 and 200 mm, respectively, as a function of etching passes of 1 to 10 using LALP at 6 W laser power and 11.4 mm/s scanning rate under water depths of 0.5 and 1.0 mm. The dash line at 500 mm depth indicates the
C.K. Chung, S.L. Lin / International Journal of Machine Tools & Manufacture 50 (2010) 961–968
963
Fig. 2. The top-view and cross-sectional optical micrographs of CO2 laser etched glass holes in air at a power of 6 W, a speed of 11.4 mm/s and 10 passes with drawn diameters of (a, b) 100 and (c, d) 200 mm. Cracks and scorches frequently appeared in this process.
through-hole etching. The etching depth increases with passes at a fixed water depth. The average etching rate can be defined by the etching depth divided by the total passes (mm/pass) and obtained from the slope of Fig. 5. It reveals that the etching rate normally decreases with increasing passes to be the non-linear behavior. The through etching of glass in both diameters of holes at the water depth of 0.5 mm occurs at around 5 passes. As the pass increases to 10, all the through holes etchings are over even at the water depth of 1.0 mm. That is, the passes of through-hole etching of glass at water depths of 0.5 and 1.0 mm are around 5 and 10, respectively. The effect of water depth on the passes of through-hole glass etching is concerned with the absorption of CO2 laser energy through water and the glass temperature cooled by water. The higher the water depth, the lower the etching depth, and the more the pass of through etching is. The drawn hole diameter will also influence the etching rate due to the variation of thermal accumulation energy. The small drawn diameter of 100 mm has lager thermal accumulation energy than that of 200 mm for higher etching depth on the same process condition. For instance, the etching depth of the hole in a drawn diameter of 100 mm at 5 scanning passes and 1.0 mm thick water is around 398 mm while that in 200 mm diameter is reduced to around 320 mm. The benefit of water adopted to assist laser through-hole etching of Pyrex glass is the cooling effect for reducing both temperature and its gradient. Therefore, the LALP method can reduce the HAZ area and eliminate the formation of cracks and scorches. Fig. 6 shows the schematic procedure of CO2 laser etched through-hole glass during LALP in order to interpret its mechanism. Because water can absorb a little CO2 laser, it is locally evaporated as laser energy goes through water layer to the surface of glass during LALP (Fig. 6(a) and (b)). Therefore, the way of laser becomes air cone, and then the glass is etched through
photothermal mechanism of melting and evaporation until through hole (Fig. 6(c)–(d)). Tending to reach the equilibrium state of liquid level, water moves to the vacant position continuously, so the laser energy is unceasingly lost by the evaporation process for reducing the etching rate. Also, the water covering around the micro-hole may reduce the temperature and its gradient for diminishing surface defect. With regard to the temperature distribution and maximum temperature during glass hole machining, a simple thermal model is adopted for the analysis of thermal distribution in air and water ambient by means of ANSYS software for solving this transient problem. It is assumed that the laser intensity distribution is a Gaussian function model and the heating phenomena due to phase changes are neglected. This simulation was performed for a laser beam power of 6 W and a scanning speed of 11.4 mm/s at an initial reference temperature of 25 1C. The laser beam focusing on the plane surface maintains a constant TEM00 mode, whose density of laser power can be described by a Gaussian distribution. The surface heat flux distribution I(x,y) can be calculated according to Eq. (1), where P0 is the laser beam power and r is the laser beam radius. The CO2 laser wavelength of 10.6 mm can be absorbed in water with a static absorption coefficient of about 500 cm 1 [25,27]. However, we cannot employ Beer’s law to calculate the quantitative values of transmitted laser intensity on glass after the initial laser intensity penetrates the water layer because the liquid-assisted laser processing is a dynamic issue including the complex photothermal interaction of water on glass related to the plasma shape, expansion and cooling along with others problems of undulation and bubble formation during laser ablation. Therefore, we introduce a correction factor to modify the absorption coefficient of CO2 laser in varied water depths. The correction factor is calculated from the ratio of experiment data in
964
C.K. Chung, S.L. Lin / International Journal of Machine Tools & Manufacture 50 (2010) 961–968
Fig. 3. (a) Top-view and (b) cross-sectional optical micrographs of the CO2 laser etched Pyrex glass material using LALP with a drawn diameter of 100 mm at a power of 6 W, a speed of 11.4 mm/s and 10 passes. The water thickness is about 0.5 mm.
Fig. 4. (a) Top-view and (b) cross-sectional optical micrographs of the CO2 laser etched Pyrex glass material using LALP with a drawn diameter of 200 mm at a power of 6 W, a speed of 11.4 mm/s and 10 passes. The water thickness is about 0.5 mm.
water to that in air: 2 p0 x þy2 Iðx,yÞ ¼ 2 exp pr r2
distance greater than 300 mm was not suitable for drilling fabrication at the above process conditions by LALP. It is because the thermal accumulation energy is not high enough to melt Pyrex glass at the center of the hole. In our larger drawn diameter test, it only etched some depth of glass but not through the wafer. The simulation result has a good agreement with our experiments. With regard to temperature gradient induced thermal stress, Fig. 7(b) shows the residual thermal stress as a function of distance from the beam center at a constant 6 W laser power and 11.4 mm/s scanning speed in air, in 0.5 and 1.0 mm thick water. The evolution of thermal stress is the same as the varied maximum temperature. The formation and propagation of crack on the laser-heated glass primarily result from the in-plane residual surface stress during ablation and cooling. The residual thermal stress (sth) is tensile and can be expressed as follows [20]:
ð1Þ
Fig. 7(a) shows the maximum temperature as a function of distance from beam center at a constant 6 W laser power and 11.4 mm/s scanning speed in air, in 0.5 and 1.0 mm thick water. The temperature decreases with increase in distance. The maximum surface temperature of glass with a drawn diameter of 100 mm can be reduced from 4009 in air to 2502 1C in 1.0 mm thick water for glass machining. When the diameter increases from 100 to 200 mm, the surface temperature of glass machining in air is reduced from 4009 to 2012 1C due to less thermal accumulation. Eq. (1) is also used to analyze the heat transfer problem for temperature distribution of glass machining in air and water ambient by ANSYS software. Table 1 lists the material property of Pyrex 7740 glass for simulation. The maximum temperature is significantly reduced in water compared to that in air. The thicker the water layer, the lower the temperature. It indicates that LALP is an effective method to avoid thermal cracking of glass through water cooling. It is noted that the
sth ð ¼ sxx ¼ syy Þ ¼
Ea ðDTÞ 1n
ð2Þ
where E, a and n are Young’s modulus, coefficient of temperature expansion (CTE) and Poisson’s ratio of Pyrex 7740 glass. DT is the increased temperature at one position during laser ablation. For
C.K. Chung, S.L. Lin / International Journal of Machine Tools & Manufacture 50 (2010) 961–968
Fig. 5. The etching depth of glass as a function of scanning pass at a fixed 6 W laser power, 0.5 and 1.0 mm water depth for 1–10 passes with a drawn diameter of (a) 100 and (b) 200 mm.
965
Fig. 7. (a) The maximum temperature and (b) the thermal stress as a function of distance from beam center at a constant 6 W laser power and 11.4 mm/s scanning speed in air, in 0.5 and 1.0 mm thick water.
Pyrex 7740 glass, E is around 63 GPa, a is around 3.25 10 6 K 1 and n is around 0.2. DT is dependent on the laser energy absorbed for glass etching. The maximum sth increases with temperature gradients, DT, and is affected by the air or water surroundings. The temperature gradient in water is smaller than that in air because of the water cooling to enhance the heat loss and reduce the overheating and thermal stress. Therefore, the thermal stress can greatly be reduced during glass etching by LALP. For example in the simulation result in Fig. 7(b), the LALP method can reduce the residual stress of glass with a drawn diameter of 100 mm from 359 MPa in air to 223 MPa in 1.0 mm thick water. Also, the Biot number (Bi) is the dimensionless number, which can be helpful to characterize the relative magnitude of the established thermal gradient [28] and defined as the ratio of the average surface heat transfer coefficient (h) for convection from the entire surface to the conductivity (k) of the solid, over a characteristic dimension L in the following equation: Bi ¼
Fig. 6. The schematic procedure of CO2 laser etched through-hole glass during LALP in order to understand its mechanism.
hL k
ð3Þ
Since h of water (500–10000 W/m2 K) is 2–3 orders higher than air (10–100 W/m2 K) [29], the large increase of Bi in water
966
C.K. Chung, S.L. Lin / International Journal of Machine Tools & Manufacture 50 (2010) 961–968
Table 1 The material property of Pyrex 7740 glass. Property
Value
Property
Value
Density (kg/m3) Thermal conduction (W/mK) Young’s modulus (Pa)
2230 1.12968 6.272 1010
Specific heat (J/kg K) Thermal expansion (1/1C) Poisson’s ratio
753.12 32.5 10 7 0.23
shows the optical micrographs of CO2 laser micromachined glass holes at the water depth of 0.5 mm at 10 10 holes array in both drawn diameters of 100 and 200 mm, respectively. The insets of pictures are the magnification of holes in arrays. Good quality of uniform through holes of Pyrex glass without cracks and scorches is also obtained using LALP technology in this easy and fast approach. Besides the cooling effect, water can be beneficial for debris removal during LALP. Fig. 9(a)–(c) shows the alpha-step profiles of through-hole glass etching with a drawn diameter of 100 mm at 6 W power and 11.4 mm/s scanning rate for 10 passes in air, under 0.5 and 1.0 mm thick water depth, respectively. The bulges at both rims of etched hole in air are around 3.75 and 7.58 mm while those by LALP at 0.5 mm thick water are around 1.15 and 1.57 mm, and at 1.0 mm thick water for 0 and 1.58 mm, respectively. The formation of bulge on the rims is mainly from resolidification of evaporated debris. Much reduced bulge height can be obtained using LALP because the laser heating induces stronger natural convection in water to help carry debris away. Assume that laser moving is along x-direction and water depth is for the z-direction. The shear stress (t) from natural convection along x-direction in cross-sectional x–z plane is defined as follows: @ux @uz þ ð4Þ t¼m @z @x
Fig. 8. The CO2 laser etched glass hole arrays using LALP at 0.5 mm thick water: (a) 10 10 arrays in 100 and (b) 10 10 arrays in 200 mm.
with pronounced heat loss leads to the decrease of both temperature itself and its gradient. Therefore, temperature gradient in water is expected to decrease much to reduce the thermal stress and HAZ areas as well to improve the crackless through-hole etching quality as shown in Figs. 3 and 4. It evidences that LALP is a very effective method for the throughhole etching of glass and has potential for application in microfluidic inlets and outlets as well as the interconnection of metal deposition for feedthrough in package. The repeatability study of through hole etching is performed in the drawn circular arrays at the efficient water depth of 0.5 mm. Fig. 8(a) and (b)
where m is the viscosity of liquid, ux and uz are the water velocities along the x-direction and z-direction, respectively. In water environment, temperature rises as the heat source passes. The temperature gradient of glass during laser ablation under water occurs because of heat accumulation on the surface of glass, which arises from the heat loss through thermal diffusion less than the heat input of laser source. The highest temperature is located at the moving laser spot on the surface and decreases with the distance away from heat source. As the laser moves, it induces natural convection with shear stress, which is proportional to water viscosity (m) and the gradient of water velocity to distance (ð@ux =@zÞ þ ð@uz =@xÞ). The lower laser moving rate or larger heat input may lead to higher shear stress. The maximum shear stress occurs at the surface of laser spot with the highest gradient of water velocity to distance. Therefore, it helps to carry away the debris under water to reduce the bulge height (Fig. 9(b) and (c)). Comparing laser ablation in air and under water, the natural convection of the former is much smaller than the later. Therefore, no effective shear force in air moves the resolidified debris away to result in the large budge (Fig. 9(a)). The vaporization for the fast etching could occur in the local laser induced hightemperature plasma environment. The glass vapor will rapidly condense due to water cooling effect. Then much of condensed glass debris can be carried away by the natural convection of water during LALP. The spot of maximum shear stress of glass under water moves with the heat source and is expected to enhance debris removal for improving surface quality. The bulge improvement result is also verified in larger diameter hole etching. Fig. 10(a)–(c) shows the alpha-step profiles of throughhole glass etching with a drawn diameter of 200 mm at 6 W power and 11.4 mm/s scanning rate for 10 passes in air under 0.5 and
C.K. Chung, S.L. Lin / International Journal of Machine Tools & Manufacture 50 (2010) 961–968
Fig. 9. The alpha-step profiles of through-hole glass etching with a drawn diameter of 100 mm at the 6 W power and 11.4 mm/s scanning rate for 10 passes: (a) in air, (b) under 0.5 thick water and (c) under 1.0 mm thick water.
1.0 mm thick water, respectively. The bulges at both rims of the etched hole in air are around 15.95 and 16.65 mm while those in 0.5 and 1.0 mm water are around 1.59–2.51 and 0.05–2.04 mm, respectively. The evolution of bulge height of glass hole with a drawn diameter of 200 mm is similar to that in smaller diameter
967
Fig. 10. The alpha-step profiles of through-hole glass etching with a drawn diameter of 200 mm at the 6 W power and 11.4 mm/s scanning rate for 10 passes: (a) in air, (b) under 0.5 thick water and (c) under 1.0 mm thick water.
of 100 mm. But the magnitude is increased due to more debris production during ablation. The thicker water depth is beneficial for debris reduction but it declines the etching rate. There is a compromise between the water depth and the etching rate for LALP performance at specific laser parameters. Overall,
968
C.K. Chung, S.L. Lin / International Journal of Machine Tools & Manufacture 50 (2010) 961–968
liquid-assisted laser processing for through-hole etching of Pyrex glass is a fast, simple and clean method at low cost.
4. Conclusion A crucial issue in Pyrex glass etching is how to etch through holes of amorphous Pyrex glass for the inlet/outlet connection or package. Here, good quality of through hole etching of Pyrex 7740 glass without cracks and scorches has been demonstrated using liquid-assisted laser processing. The cracks or scorches are formed using CO2 laser etched glass in air due to the high temperature induced thermal stress. The water is used to reduce the temperature and its gradient to avoid the formation of cracks or scorches. A through hole of Pyrex glass is easy to obtain using LALP to etch a circular shape for 5–10 passes at a 6 W power and 11.4 mm/s scanning speed via the drawn circular diameter of 100–200 mm in several seconds. The hole arrays are also obtained to prove the repeatability of through hole etching of Pyrex glass. LALP technology has advantages of high etching rate, easy fabrication, and low cost with no crack and scorches. ANSYS software was employed to analyze the temperature distribution and thermal stress field. The higher temperature gradient in air induced higher stress for crack formation while the smaller temperature gradient in water had the less heat-affected zone range and eliminated the crack during processing. The simulation result is in a good agreement with the experiments.
Acknowledgements This work was partly sponsored by National Science Council (NSC) under contract No NSC 95-2221-E006-047-MY3 and 982221-E-006-052-MY3. We are greatly indebted to the Center for Micro/Nano Science and Technology (CMNST) at National Cheng Kung University, National Nano Device Laboratories (NDL) and National Center for High-performing Computing (NCHC) for the access of process, computation and analysis equipments. References [1] T. Cormany, P. Enoksson, G. Stemme, Deep wet etching of borosilicate glass using an anodically bonded silicon substrate as mask, J. Micromech. Microeng. 8 (1998) 84–87. [2] D.C.S. Bien, P.V. Rainey, S.J.N. Mitchell, H.S. Gamble, Characterization of masking materials for deep glass micromachining, J. Micromech. Microeng. 13 (2003) S34–S40. [3] C. Iliescu, J. Miao, F.E.H. Taya, Stress control in masking layers for deep wet micromachining of Pyrex glass, Sensors Actuators A 117 (2005) 286–292. [4] X. Li, T. Abe, M. Esashi, Deep reactive ion etching of Pyrex glass using SF6 plasma, Sensors Actuators A 87 (2001) 139–145.
[5] D.J. Kim, Y. Ahn, S.H. Lee, Y.K. Kim, Voltage pulse frequency and duty ratio effects in an electrochemical discharge microdrilling process of Pyrex glass, Int. J. Mach. Tools Manuf. 46 (2006) 1064–1067. [6] C.S. Nielsen, P. Balling, Deep drilling of metals with ultrashort laser pulses: a two-stage process, J. Appl. Phys. 99 (2006) 093101. [7] A. Weck, T.H.R. Crawford, D.S. Wilkinson, H.K. Haugen, J.S. Preston, Laser drilling of high aspect ratio holes in copper with femtosecond, picosecond and nanosecond pulse, Appl. Phys. A 90 (2008) 537–543. [8] J. Cheng, W. Perrie, M. Sharp, S.P. Edwardson, N.G. Semaltianos, G. Dearden, K.G. Watkins, Single-pulse drilling study on Au, AL and Ti alloy by using a picosecond laser, Appl. Phys. A 95 (2009) 739–746. [9] K. Venkatakrishnan, N. Sudani, B. Tan, A high-repetition-rate femtosecond laser for thin silicon wafer dicing, J. Micromech. Microeng. 18 (2008) 075032 (7pp). [10] B. Tan, K. Venkatakrishnan, Nd–YAG laser microvia drilling for interconnection application, J. Micromech. Microeng. 17 (2007) 1511–1517. [11] B. Tan, S. Panchatsharam, K. Venkatakrishnan, High repetition rate femtosecond laser forming sub-10 mm diameter interconnection vias, J. Phys. D: Appl. Phys. 42 (2009) 065102 (9pp.). [12] H.Y. Zheng, Y.C. Lam, C. Sundarraman, D.V. Tran, Influence of substrate cooling on femtosecond laser machined hole depth and diameter, Appl. Phys. A 89 (2007) 559–563. [13] R. Petkovˇsek, A. Babnik, J. Diaci, J. Mozˇina, Optodynamic monitoring of laser micro-drilling of glass by using a laser probe, Appl. Phys. A 93 (2008) 141–145. [14] Q. Sun, A. Saliminia, F. The´berge, R. Valle´e, S.L. Chin, Microchannel fabrication in silica glass by femtosecond laser pulses with different central wavelength, J. Micromech. Microeng. 18 (2008) 035039 (4pp). [15] R. Bohme, K. Zimmer, B. Rauschenbach, Laser backside etching of fused silica due to carbon layer ablation, Appl. Phys. A 82 (2006) 325–328. [16] J.Y. Cheng, M.H. Yen, C.W. Wei, Y.C. Chuang, T.H. Young, Crack-free directwriting on glass using a low-power UV laser in the manufacture of a microfluidic chip, J. Micromech. Microeng. 15 (2005) 1147–1156. [17] A. Kruusing, Underwater and water-assisted laser processing: Part 2—etching, cutting and rarely used methods, Opt. Lasers Eng. 41 (2004) 329–352. [18] R. An, M.D. Hoffman, M.A. Donoghue, A.J. Hunt, S.C. Jacobson, Water-assisted femtosecond laser machining of electrospray nozzles on glass microfluidic devices, Opt Express 16 (2008) 15206–15211. [19] C.K. Chung, Y.C. Lin, G.R. Huang, Bulge formation and improvement of the polymer in CO2 laser micromachining, J. Micromech. Microeng. 15 (2005) 1878–1884. [20] G. Allcock, P.E. Dyer, G. Elliner, H.V. Snelling, Experimental observations and analysis of CO2 laser-induced microcracking of glass, J. Appl. Phys. 78 (1995) 7295–7303. [21] H. Ogura, Y. Yoshida, Hole drilling of glass substrates with a CO2 laser, Jpn. J. Appl. Phys. 42 (2003) 2881–2886. [22] M.H. Yen, J.Y. Cheng, C.W. Wei, Y.C. Chung, T.H. Young, Rapid cell-patterning and microfluidic chip fabrication by crack-free CO2 laser ablation on glass, J. Micromech. Microeng. 16 (2006) 1143–1153. [23] C.K. Chung, Y.C. Sung, G.R. Huang, E.J. Hsiao, W.H. Lin, S.L. Lin, Crackless linear through-wafer etching of Pyrex glass using liquid-assisted CO2 laser processing, Appl. Phys. A 94 (2009) 927–932. [24] C. Barnes, P. Shrotriya, P. Molian, Water-assisted laser thermal shock machining of alumina, Int. J. Mach. Tools Manuf. 47 (2007) 1864–1874. [25] C.K. Chung, H.C. Chang, T.R. Shih, S.L. Lin, E.J. Hsiao, Y.S. Chen, E.C. Chang, C.C. Chen, C.C. Lin, Water-assisted CO2 laser ablated glass and modified thermal bonding for capillary-driven bio-fluidic application, Biomed. Microdevices 12 (2010) 107–114. [26] C.K. Chung, M.Y. Wu, A hybrid CO2 laser processing for silicon etching, Opt. Express 15 (2007) 7269–7274. [27] G.M. Hale, M.R. Querry, Optical constants of water in the 200-nm to 200-mm wavelength region, Appl. Opt. 12 (1973) 555–563. [28] Marc Madou, in: Fundamentals of Microfabrication, second ed., CRC press, New York, 2002, p. 590. [29] /http://www.engineeringtoolbox.com/overall-heat-transfer-coefficient-d_434. htmlS.
Contents lists available at ScienceDirect
International Journal of Machine Tools & Manufacture journal homepage: www.elsevier.com/locate/ijmactool
CO2 laser micromachined crackless through holes of Pyrex 7740 glass C.K. Chung n, S.L. Lin Department of Mechanical Engineering, and Center for Micro/Nano Science and Technology, National Cheng Kung University, Tainan, Taiwan
a r t i c l e in fo
abstract
Article history: Received 19 February 2010 Received in revised form 8 August 2010 Accepted 9 August 2010 Available online 13 August 2010
The drilling of glass through holes with a high aspect ratio is crucial for microsystems application, especially in the inlet/outlet connection of microfluidic devices for biological analysis or for the anodic bonded silicon-glass ones. Traditional glass drilling using mechanical processing and laser processing in air would produce many kinds of defects such as bulges, debris, cracks and scorch. In this paper, we have applied the method of liquid-assisted laser processing (LALP) to reduce the temperature gradient, bulges and heat affected zone (HAZ) region for achieving crack-free glass machined holes. The nominal diameters of circles from 100 to 200 mm were drawn for through glass machining test. Through-hole glass etching can be obtained by LALP for 10 passes of circular scanning in several seconds on conditions of a 6 W laser power, 76 mm spot size and 11.4 mm/s scanning speed. The ANSYS software was also used to analyze the temperature distribution and thermal stress field in air and water ambient during glass hole machining. The higher temperature gradient in air induced higher stress for crack formation while the smaller temperature gradient in water had less HAZ and eliminated the crack during processing. CO2 laser micromachining under water has merits of high etching rate, easy fabrication and low cost together with much improved surface quality compared to that in air. & 2010 Elsevier Ltd. All rights reserved.
Keywords: Pyrex glass CO2 laser Water-assisted Etching Through hole
1. Introduction Glass materials have widely been used in the application of optical and optoelectric devices and microfluidic and bio MEMS. Glass micromachining, especially deep etching of Pyrex glass, was studied by many researchers using wet hydrofluoric acid (HF) based solution [1–3], dry plasma etching [4] or electrochemical discharge microdrilling [5]. HF based solution with the assistance of photolithographic masking can selectively etch glass up to 500 mm deep but it encounters problems of large lateral undercut and very low etching rate of about 12.9 mm/h [1]. The masking materials with the problem of pinhole formation often existed during deep glass etching [2]. The annealed PECVD amorphous Si among different masking materials of Cr/Au, Cr/Cu and itself was tested to be the best one for glass etching up to 200 mm in HF (49%) solution [3]. Deep etching of Pyrex glass using SF6 plasma in a high-cost ICP system also exhibited a low etching rate of about 0.6 mm/min [4]. Electrochemical discharge microdrilling had one drawback: heat-affected zone (HAZ) being left on the microdrilled hole surface [5]. Recently, laser has widely been studied in micromachining. Ultra-short laser systems, picosecond to femtosecond lasers, were used for the fabrication of drilled holes and shaped microstructure in metal materials [6–8] or silicon wafers [9–11]. The ultra-short laser has the advantage of a small heat
n
Corresponding author. Tel.: 886 6 2757575x62111; fax: 886 6 2352973. E-mail address: [email protected] (C.K. Chung).
0890-6955/$ - see front matter & 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.ijmachtools.2010.08.002
affected zone (HAZ) on the surrounding material due to the minimized heat conduction during a short pulse. Several publications used the assisted method to etch transparent silica glass at low repetition rate or higher output laser energy [12–14]. But the expensive femtosecond laser or excimer laser [15] used for etching glass had low etching rates together with the high cost, especially for deep through-wafer etching. The traditional defects of bulge, debris, cracks and scorch often occurred during laser processing to influence the fabrication quality for application. Cheng et al. [16] reported acid solution assisted UV-laser wet etching glass substrate for microfluidic application but the etching rate was still low, together with the occurrence of traditional high-bulge defects. A liquid-assisted laser processing (LALP) has been applied to the welding, cleaning or cutting of metals, alloys and crystalline ceramics [17,18]. The drilling of through holes with a high aspect ratio is important for potential industrial application. Although CO2 laser was widely used for transparent material ablation [19], it was rarely used to directly etch through amorphous Pyrex glass at room temperature in air due to the thermal stress induced crack problem [20]. Ogura and Yoshida [21] reported crystalline and amorphous glass hole drilling using CO2 laser and found that crystalline synthetic quartz had the best drilling quality together with through-hole drilling while amorphous Pyrex glass had the worst result together with the highest bulge around the hole as well as no through-hole cross section being demonstrated. Yen et al. [22] reported the substrate heating method to improve the defects of CO2 laser ablation on glass for microfluidic application. But no
962
C.K. Chung, S.L. Lin / International Journal of Machine Tools & Manufacture 50 (2010) 961–968
through-hole etching was shown and such method would affect the cutting performance and increase instrument complexity. Chung et al. [23] reported the linearly through-wafer etching of Pyrex glass using liquid-assisted CO2 laser process but with no through-hole etching. Barnes et al. [24] proposed the water assisted CO2 laser cutting alumina specimen compared with laser processing in air, then a finite element approach was used to analyze the temperature and thermal stress distributions during both cuttings. Chung et al. [25] used the combination of waterassisted CO2 laser processing for ablating glass slides and the modified thermal bonding for capillary-driven bio-fluidic application. In this paper, we have demonstrated the effective method of through-hole etching of amorphous Pyrex glass using the lowcost CO2 laser based on LALP technology. The liquid form of water was used to obtain good quality of through hole glass etching for the sake of cooling and cleaning. In order to further investigate the temperature distribution, maximum temperature and thermal stress during the drilling machining process, a simple thermal model was adopted for the analysis of thermal behavior during machining. ANSYS software was used to solve this transient problem. The water cooling effect and debris removal mechanism were also discussed for drilling quality improvement.
power of 6 W and scanning speed of 11.4 mm/s at different scanning numbers from 1 to 10 passes were applied to glass etching. Both kinds of water depth were controlled at about 0.5 and 1.0 mm above glass surface by the total volume method whose depth is proportional to the volume at a constant area of holder. In order to keep the glass surface always wet during ablation, 5 ml of hydrophilic surfactant agent of lauramidopropyl betaine (RCONH(CH2)3N + (CH3)2CH2COO–) was added to avoid the water layer broken into drops. Traditional laser processing in air was performed as a reference sample. The optical microscope (OLYMPUS BX51M, Japan) was used to examine the morphology after laser processing. The profile and morphology of lasermachined glass channels were examined by the alpha-step (a-step) profiler (Kosaka Lab, ET3000, Japan). ANSYS software was used to investigate the temperature distribution and thermal stress field in air and water ambient during glass hole machining. The laser beam focusing on the plane surface maintains a constant TEM00 mode, and laser intensity distribution is a Gaussian function model and the heating phenomena due to phase changes are neglected. The beam of CO2 laser is regarded as a surface heating source but the surface of Pyrex 7740 glass without laser heating and the superficial heat irradiation is negligible.
3. Results and discussion 2. Fabrication and experimental procedures Fig. 1 shows the schematic diagram of liquid-assisted CO2 laser processing setup. It includes the CO2 laser source, a reflective mirror, a focal lens and a sample immersed in water. The test sample is amorphous Pyrex 7740 glass with an average thickness of 500 mm. The glass was submerged in water in a container at room temperature, and then laser beam was performed to etch glass at proper parameters. The working liquid selected was water due to its chemical inertness. The commercial available air-cooled CO2 laser equipment (VL-200, Universal Laser system Inc., USA) was used with a maximum laser power of 30 W. The CO2 laser has a wavelength of 10.6 mm and a TEM00 output of beam. Also, the CO2 laser uses a sealed-off, RF excited, slab design and a multipass, free space resonator for a good quality beam with an M2 value of 1.4 70.2. The maximum laser scanning speed was 1140 mm/s and the largest working area was 409 304 mm2. The focal length of the lens is about 38.1 mm and the smallest beam spot size of the commercial product after standard verification could reach 76 mm, which is defined as the double distance from the center of Gaussian intensity distribution to that with 1/e2 maximum intensity. The laser spot was focused on the top surface of Pyrex glass during LALP. The laser power, scanning speed and the number of drilling passes were controlled to get a better processing result. A computer aided design program of CorelDraw software was available to set the experimental parameters for auto-controlled processing. The constant laser
Fig. 1. Schematic diagram of CO2 laser micromachining setup. It includes the CO2 laser source, a reflective mirror, a focal lens and an experimental sample immersed in water.
Fig. 2(a)–(d) shows the top-view and cross-sectional optical micrographs of CO2 laser etched glass holes in air with drawn diameters of 100 (Fig. 2(a)–(b)) and 200 mm (Fig. 2(c)–(d)), respectively, at a power of 6 W, a speed of 11.4 mm/s and 10 passes. Cracks and scorches frequently appeared on the glass surface during laser processing in air. The cracks occur prior to through-hole glass etching because of the high thermal stress [20] while the scorch formation is mainly from the huge amount of accumulated heating energy for burning, especially more in circular etching than in linear cutting [26]. Since the focal length of the lens is about 38.1 mm and the laser beam intensity is a Gaussian distribution, the etch rate is higher at the center than that at the rim of hole. In addition, a clear HAZ area can be found in Fig. 2(b) and (d) due to the existence of temperature gradient. The higher temperature gradient leads to the larger HAZ area even cracks because the drawn diameter of 100 mm has larger thermal accumulation energy than that of 200 mm. The CO2 laser micromachined glass hole using LALP on a drawn diameter of 100 mm is shown in Fig. 3(a) and (b) in terms of top view and cross section. The liquid is water and about 0.5 mm above the surface of glass. The laser operation parameters are the same as those in air in Fig. 2(a) and (b). No cracks and scorches appear. The HAZ area by LALP is smaller than that in air. The central white spot indicates the through-hole etching and it has a smaller diameter than the total circle due to the tapered profile (Fig. 3(b)) formed by laser ablation. Similar results can be obtained by etching of a hole in larger diameter. Fig. 4(a) and (b) shows the top-view and cross-sectional optical micrographs of the CO2 laser etched Pyrex glass with a drawn diameter of 200 mm using LALP, respectively, at a power of 6 W, a speed of 11.4 mm/s and 10 passes. The water depth is the same 0.5 mm above the surface of glass. Also no crack or scorch is formed under LALP. The tapered profile is more diffusive than that in the small diameter hole. The water depth will affect the efficient laser power and etching rate. Fig. 5(a) and (b) shows the etching depths of the holes in drawn diameters of 100 and 200 mm, respectively, as a function of etching passes of 1 to 10 using LALP at 6 W laser power and 11.4 mm/s scanning rate under water depths of 0.5 and 1.0 mm. The dash line at 500 mm depth indicates the
C.K. Chung, S.L. Lin / International Journal of Machine Tools & Manufacture 50 (2010) 961–968
963
Fig. 2. The top-view and cross-sectional optical micrographs of CO2 laser etched glass holes in air at a power of 6 W, a speed of 11.4 mm/s and 10 passes with drawn diameters of (a, b) 100 and (c, d) 200 mm. Cracks and scorches frequently appeared in this process.
through-hole etching. The etching depth increases with passes at a fixed water depth. The average etching rate can be defined by the etching depth divided by the total passes (mm/pass) and obtained from the slope of Fig. 5. It reveals that the etching rate normally decreases with increasing passes to be the non-linear behavior. The through etching of glass in both diameters of holes at the water depth of 0.5 mm occurs at around 5 passes. As the pass increases to 10, all the through holes etchings are over even at the water depth of 1.0 mm. That is, the passes of through-hole etching of glass at water depths of 0.5 and 1.0 mm are around 5 and 10, respectively. The effect of water depth on the passes of through-hole glass etching is concerned with the absorption of CO2 laser energy through water and the glass temperature cooled by water. The higher the water depth, the lower the etching depth, and the more the pass of through etching is. The drawn hole diameter will also influence the etching rate due to the variation of thermal accumulation energy. The small drawn diameter of 100 mm has lager thermal accumulation energy than that of 200 mm for higher etching depth on the same process condition. For instance, the etching depth of the hole in a drawn diameter of 100 mm at 5 scanning passes and 1.0 mm thick water is around 398 mm while that in 200 mm diameter is reduced to around 320 mm. The benefit of water adopted to assist laser through-hole etching of Pyrex glass is the cooling effect for reducing both temperature and its gradient. Therefore, the LALP method can reduce the HAZ area and eliminate the formation of cracks and scorches. Fig. 6 shows the schematic procedure of CO2 laser etched through-hole glass during LALP in order to interpret its mechanism. Because water can absorb a little CO2 laser, it is locally evaporated as laser energy goes through water layer to the surface of glass during LALP (Fig. 6(a) and (b)). Therefore, the way of laser becomes air cone, and then the glass is etched through
photothermal mechanism of melting and evaporation until through hole (Fig. 6(c)–(d)). Tending to reach the equilibrium state of liquid level, water moves to the vacant position continuously, so the laser energy is unceasingly lost by the evaporation process for reducing the etching rate. Also, the water covering around the micro-hole may reduce the temperature and its gradient for diminishing surface defect. With regard to the temperature distribution and maximum temperature during glass hole machining, a simple thermal model is adopted for the analysis of thermal distribution in air and water ambient by means of ANSYS software for solving this transient problem. It is assumed that the laser intensity distribution is a Gaussian function model and the heating phenomena due to phase changes are neglected. This simulation was performed for a laser beam power of 6 W and a scanning speed of 11.4 mm/s at an initial reference temperature of 25 1C. The laser beam focusing on the plane surface maintains a constant TEM00 mode, whose density of laser power can be described by a Gaussian distribution. The surface heat flux distribution I(x,y) can be calculated according to Eq. (1), where P0 is the laser beam power and r is the laser beam radius. The CO2 laser wavelength of 10.6 mm can be absorbed in water with a static absorption coefficient of about 500 cm 1 [25,27]. However, we cannot employ Beer’s law to calculate the quantitative values of transmitted laser intensity on glass after the initial laser intensity penetrates the water layer because the liquid-assisted laser processing is a dynamic issue including the complex photothermal interaction of water on glass related to the plasma shape, expansion and cooling along with others problems of undulation and bubble formation during laser ablation. Therefore, we introduce a correction factor to modify the absorption coefficient of CO2 laser in varied water depths. The correction factor is calculated from the ratio of experiment data in
964
C.K. Chung, S.L. Lin / International Journal of Machine Tools & Manufacture 50 (2010) 961–968
Fig. 3. (a) Top-view and (b) cross-sectional optical micrographs of the CO2 laser etched Pyrex glass material using LALP with a drawn diameter of 100 mm at a power of 6 W, a speed of 11.4 mm/s and 10 passes. The water thickness is about 0.5 mm.
Fig. 4. (a) Top-view and (b) cross-sectional optical micrographs of the CO2 laser etched Pyrex glass material using LALP with a drawn diameter of 200 mm at a power of 6 W, a speed of 11.4 mm/s and 10 passes. The water thickness is about 0.5 mm.
water to that in air: 2 p0 x þy2 Iðx,yÞ ¼ 2 exp pr r2
distance greater than 300 mm was not suitable for drilling fabrication at the above process conditions by LALP. It is because the thermal accumulation energy is not high enough to melt Pyrex glass at the center of the hole. In our larger drawn diameter test, it only etched some depth of glass but not through the wafer. The simulation result has a good agreement with our experiments. With regard to temperature gradient induced thermal stress, Fig. 7(b) shows the residual thermal stress as a function of distance from the beam center at a constant 6 W laser power and 11.4 mm/s scanning speed in air, in 0.5 and 1.0 mm thick water. The evolution of thermal stress is the same as the varied maximum temperature. The formation and propagation of crack on the laser-heated glass primarily result from the in-plane residual surface stress during ablation and cooling. The residual thermal stress (sth) is tensile and can be expressed as follows [20]:
ð1Þ
Fig. 7(a) shows the maximum temperature as a function of distance from beam center at a constant 6 W laser power and 11.4 mm/s scanning speed in air, in 0.5 and 1.0 mm thick water. The temperature decreases with increase in distance. The maximum surface temperature of glass with a drawn diameter of 100 mm can be reduced from 4009 in air to 2502 1C in 1.0 mm thick water for glass machining. When the diameter increases from 100 to 200 mm, the surface temperature of glass machining in air is reduced from 4009 to 2012 1C due to less thermal accumulation. Eq. (1) is also used to analyze the heat transfer problem for temperature distribution of glass machining in air and water ambient by ANSYS software. Table 1 lists the material property of Pyrex 7740 glass for simulation. The maximum temperature is significantly reduced in water compared to that in air. The thicker the water layer, the lower the temperature. It indicates that LALP is an effective method to avoid thermal cracking of glass through water cooling. It is noted that the
sth ð ¼ sxx ¼ syy Þ ¼
Ea ðDTÞ 1n
ð2Þ
where E, a and n are Young’s modulus, coefficient of temperature expansion (CTE) and Poisson’s ratio of Pyrex 7740 glass. DT is the increased temperature at one position during laser ablation. For
C.K. Chung, S.L. Lin / International Journal of Machine Tools & Manufacture 50 (2010) 961–968
Fig. 5. The etching depth of glass as a function of scanning pass at a fixed 6 W laser power, 0.5 and 1.0 mm water depth for 1–10 passes with a drawn diameter of (a) 100 and (b) 200 mm.
965
Fig. 7. (a) The maximum temperature and (b) the thermal stress as a function of distance from beam center at a constant 6 W laser power and 11.4 mm/s scanning speed in air, in 0.5 and 1.0 mm thick water.
Pyrex 7740 glass, E is around 63 GPa, a is around 3.25 10 6 K 1 and n is around 0.2. DT is dependent on the laser energy absorbed for glass etching. The maximum sth increases with temperature gradients, DT, and is affected by the air or water surroundings. The temperature gradient in water is smaller than that in air because of the water cooling to enhance the heat loss and reduce the overheating and thermal stress. Therefore, the thermal stress can greatly be reduced during glass etching by LALP. For example in the simulation result in Fig. 7(b), the LALP method can reduce the residual stress of glass with a drawn diameter of 100 mm from 359 MPa in air to 223 MPa in 1.0 mm thick water. Also, the Biot number (Bi) is the dimensionless number, which can be helpful to characterize the relative magnitude of the established thermal gradient [28] and defined as the ratio of the average surface heat transfer coefficient (h) for convection from the entire surface to the conductivity (k) of the solid, over a characteristic dimension L in the following equation: Bi ¼
Fig. 6. The schematic procedure of CO2 laser etched through-hole glass during LALP in order to understand its mechanism.
hL k
ð3Þ
Since h of water (500–10000 W/m2 K) is 2–3 orders higher than air (10–100 W/m2 K) [29], the large increase of Bi in water
966
C.K. Chung, S.L. Lin / International Journal of Machine Tools & Manufacture 50 (2010) 961–968
Table 1 The material property of Pyrex 7740 glass. Property
Value
Property
Value
Density (kg/m3) Thermal conduction (W/mK) Young’s modulus (Pa)
2230 1.12968 6.272 1010
Specific heat (J/kg K) Thermal expansion (1/1C) Poisson’s ratio
753.12 32.5 10 7 0.23
shows the optical micrographs of CO2 laser micromachined glass holes at the water depth of 0.5 mm at 10 10 holes array in both drawn diameters of 100 and 200 mm, respectively. The insets of pictures are the magnification of holes in arrays. Good quality of uniform through holes of Pyrex glass without cracks and scorches is also obtained using LALP technology in this easy and fast approach. Besides the cooling effect, water can be beneficial for debris removal during LALP. Fig. 9(a)–(c) shows the alpha-step profiles of through-hole glass etching with a drawn diameter of 100 mm at 6 W power and 11.4 mm/s scanning rate for 10 passes in air, under 0.5 and 1.0 mm thick water depth, respectively. The bulges at both rims of etched hole in air are around 3.75 and 7.58 mm while those by LALP at 0.5 mm thick water are around 1.15 and 1.57 mm, and at 1.0 mm thick water for 0 and 1.58 mm, respectively. The formation of bulge on the rims is mainly from resolidification of evaporated debris. Much reduced bulge height can be obtained using LALP because the laser heating induces stronger natural convection in water to help carry debris away. Assume that laser moving is along x-direction and water depth is for the z-direction. The shear stress (t) from natural convection along x-direction in cross-sectional x–z plane is defined as follows: @ux @uz þ ð4Þ t¼m @z @x
Fig. 8. The CO2 laser etched glass hole arrays using LALP at 0.5 mm thick water: (a) 10 10 arrays in 100 and (b) 10 10 arrays in 200 mm.
with pronounced heat loss leads to the decrease of both temperature itself and its gradient. Therefore, temperature gradient in water is expected to decrease much to reduce the thermal stress and HAZ areas as well to improve the crackless through-hole etching quality as shown in Figs. 3 and 4. It evidences that LALP is a very effective method for the throughhole etching of glass and has potential for application in microfluidic inlets and outlets as well as the interconnection of metal deposition for feedthrough in package. The repeatability study of through hole etching is performed in the drawn circular arrays at the efficient water depth of 0.5 mm. Fig. 8(a) and (b)
where m is the viscosity of liquid, ux and uz are the water velocities along the x-direction and z-direction, respectively. In water environment, temperature rises as the heat source passes. The temperature gradient of glass during laser ablation under water occurs because of heat accumulation on the surface of glass, which arises from the heat loss through thermal diffusion less than the heat input of laser source. The highest temperature is located at the moving laser spot on the surface and decreases with the distance away from heat source. As the laser moves, it induces natural convection with shear stress, which is proportional to water viscosity (m) and the gradient of water velocity to distance (ð@ux =@zÞ þ ð@uz =@xÞ). The lower laser moving rate or larger heat input may lead to higher shear stress. The maximum shear stress occurs at the surface of laser spot with the highest gradient of water velocity to distance. Therefore, it helps to carry away the debris under water to reduce the bulge height (Fig. 9(b) and (c)). Comparing laser ablation in air and under water, the natural convection of the former is much smaller than the later. Therefore, no effective shear force in air moves the resolidified debris away to result in the large budge (Fig. 9(a)). The vaporization for the fast etching could occur in the local laser induced hightemperature plasma environment. The glass vapor will rapidly condense due to water cooling effect. Then much of condensed glass debris can be carried away by the natural convection of water during LALP. The spot of maximum shear stress of glass under water moves with the heat source and is expected to enhance debris removal for improving surface quality. The bulge improvement result is also verified in larger diameter hole etching. Fig. 10(a)–(c) shows the alpha-step profiles of throughhole glass etching with a drawn diameter of 200 mm at 6 W power and 11.4 mm/s scanning rate for 10 passes in air under 0.5 and
C.K. Chung, S.L. Lin / International Journal of Machine Tools & Manufacture 50 (2010) 961–968
Fig. 9. The alpha-step profiles of through-hole glass etching with a drawn diameter of 100 mm at the 6 W power and 11.4 mm/s scanning rate for 10 passes: (a) in air, (b) under 0.5 thick water and (c) under 1.0 mm thick water.
1.0 mm thick water, respectively. The bulges at both rims of the etched hole in air are around 15.95 and 16.65 mm while those in 0.5 and 1.0 mm water are around 1.59–2.51 and 0.05–2.04 mm, respectively. The evolution of bulge height of glass hole with a drawn diameter of 200 mm is similar to that in smaller diameter
967
Fig. 10. The alpha-step profiles of through-hole glass etching with a drawn diameter of 200 mm at the 6 W power and 11.4 mm/s scanning rate for 10 passes: (a) in air, (b) under 0.5 thick water and (c) under 1.0 mm thick water.
of 100 mm. But the magnitude is increased due to more debris production during ablation. The thicker water depth is beneficial for debris reduction but it declines the etching rate. There is a compromise between the water depth and the etching rate for LALP performance at specific laser parameters. Overall,
968
C.K. Chung, S.L. Lin / International Journal of Machine Tools & Manufacture 50 (2010) 961–968
liquid-assisted laser processing for through-hole etching of Pyrex glass is a fast, simple and clean method at low cost.
4. Conclusion A crucial issue in Pyrex glass etching is how to etch through holes of amorphous Pyrex glass for the inlet/outlet connection or package. Here, good quality of through hole etching of Pyrex 7740 glass without cracks and scorches has been demonstrated using liquid-assisted laser processing. The cracks or scorches are formed using CO2 laser etched glass in air due to the high temperature induced thermal stress. The water is used to reduce the temperature and its gradient to avoid the formation of cracks or scorches. A through hole of Pyrex glass is easy to obtain using LALP to etch a circular shape for 5–10 passes at a 6 W power and 11.4 mm/s scanning speed via the drawn circular diameter of 100–200 mm in several seconds. The hole arrays are also obtained to prove the repeatability of through hole etching of Pyrex glass. LALP technology has advantages of high etching rate, easy fabrication, and low cost with no crack and scorches. ANSYS software was employed to analyze the temperature distribution and thermal stress field. The higher temperature gradient in air induced higher stress for crack formation while the smaller temperature gradient in water had the less heat-affected zone range and eliminated the crack during processing. The simulation result is in a good agreement with the experiments.
Acknowledgements This work was partly sponsored by National Science Council (NSC) under contract No NSC 95-2221-E006-047-MY3 and 982221-E-006-052-MY3. We are greatly indebted to the Center for Micro/Nano Science and Technology (CMNST) at National Cheng Kung University, National Nano Device Laboratories (NDL) and National Center for High-performing Computing (NCHC) for the access of process, computation and analysis equipments. References [1] T. Cormany, P. Enoksson, G. Stemme, Deep wet etching of borosilicate glass using an anodically bonded silicon substrate as mask, J. Micromech. Microeng. 8 (1998) 84–87. [2] D.C.S. Bien, P.V. Rainey, S.J.N. Mitchell, H.S. Gamble, Characterization of masking materials for deep glass micromachining, J. Micromech. Microeng. 13 (2003) S34–S40. [3] C. Iliescu, J. Miao, F.E.H. Taya, Stress control in masking layers for deep wet micromachining of Pyrex glass, Sensors Actuators A 117 (2005) 286–292. [4] X. Li, T. Abe, M. Esashi, Deep reactive ion etching of Pyrex glass using SF6 plasma, Sensors Actuators A 87 (2001) 139–145.
[5] D.J. Kim, Y. Ahn, S.H. Lee, Y.K. Kim, Voltage pulse frequency and duty ratio effects in an electrochemical discharge microdrilling process of Pyrex glass, Int. J. Mach. Tools Manuf. 46 (2006) 1064–1067. [6] C.S. Nielsen, P. Balling, Deep drilling of metals with ultrashort laser pulses: a two-stage process, J. Appl. Phys. 99 (2006) 093101. [7] A. Weck, T.H.R. Crawford, D.S. Wilkinson, H.K. Haugen, J.S. Preston, Laser drilling of high aspect ratio holes in copper with femtosecond, picosecond and nanosecond pulse, Appl. Phys. A 90 (2008) 537–543. [8] J. Cheng, W. Perrie, M. Sharp, S.P. Edwardson, N.G. Semaltianos, G. Dearden, K.G. Watkins, Single-pulse drilling study on Au, AL and Ti alloy by using a picosecond laser, Appl. Phys. A 95 (2009) 739–746. [9] K. Venkatakrishnan, N. Sudani, B. Tan, A high-repetition-rate femtosecond laser for thin silicon wafer dicing, J. Micromech. Microeng. 18 (2008) 075032 (7pp). [10] B. Tan, K. Venkatakrishnan, Nd–YAG laser microvia drilling for interconnection application, J. Micromech. Microeng. 17 (2007) 1511–1517. [11] B. Tan, S. Panchatsharam, K. Venkatakrishnan, High repetition rate femtosecond laser forming sub-10 mm diameter interconnection vias, J. Phys. D: Appl. Phys. 42 (2009) 065102 (9pp.). [12] H.Y. Zheng, Y.C. Lam, C. Sundarraman, D.V. Tran, Influence of substrate cooling on femtosecond laser machined hole depth and diameter, Appl. Phys. A 89 (2007) 559–563. [13] R. Petkovˇsek, A. Babnik, J. Diaci, J. Mozˇina, Optodynamic monitoring of laser micro-drilling of glass by using a laser probe, Appl. Phys. A 93 (2008) 141–145. [14] Q. Sun, A. Saliminia, F. The´berge, R. Valle´e, S.L. Chin, Microchannel fabrication in silica glass by femtosecond laser pulses with different central wavelength, J. Micromech. Microeng. 18 (2008) 035039 (4pp). [15] R. Bohme, K. Zimmer, B. Rauschenbach, Laser backside etching of fused silica due to carbon layer ablation, Appl. Phys. A 82 (2006) 325–328. [16] J.Y. Cheng, M.H. Yen, C.W. Wei, Y.C. Chuang, T.H. Young, Crack-free directwriting on glass using a low-power UV laser in the manufacture of a microfluidic chip, J. Micromech. Microeng. 15 (2005) 1147–1156. [17] A. Kruusing, Underwater and water-assisted laser processing: Part 2—etching, cutting and rarely used methods, Opt. Lasers Eng. 41 (2004) 329–352. [18] R. An, M.D. Hoffman, M.A. Donoghue, A.J. Hunt, S.C. Jacobson, Water-assisted femtosecond laser machining of electrospray nozzles on glass microfluidic devices, Opt Express 16 (2008) 15206–15211. [19] C.K. Chung, Y.C. Lin, G.R. Huang, Bulge formation and improvement of the polymer in CO2 laser micromachining, J. Micromech. Microeng. 15 (2005) 1878–1884. [20] G. Allcock, P.E. Dyer, G. Elliner, H.V. Snelling, Experimental observations and analysis of CO2 laser-induced microcracking of glass, J. Appl. Phys. 78 (1995) 7295–7303. [21] H. Ogura, Y. Yoshida, Hole drilling of glass substrates with a CO2 laser, Jpn. J. Appl. Phys. 42 (2003) 2881–2886. [22] M.H. Yen, J.Y. Cheng, C.W. Wei, Y.C. Chung, T.H. Young, Rapid cell-patterning and microfluidic chip fabrication by crack-free CO2 laser ablation on glass, J. Micromech. Microeng. 16 (2006) 1143–1153. [23] C.K. Chung, Y.C. Sung, G.R. Huang, E.J. Hsiao, W.H. Lin, S.L. Lin, Crackless linear through-wafer etching of Pyrex glass using liquid-assisted CO2 laser processing, Appl. Phys. A 94 (2009) 927–932. [24] C. Barnes, P. Shrotriya, P. Molian, Water-assisted laser thermal shock machining of alumina, Int. J. Mach. Tools Manuf. 47 (2007) 1864–1874. [25] C.K. Chung, H.C. Chang, T.R. Shih, S.L. Lin, E.J. Hsiao, Y.S. Chen, E.C. Chang, C.C. Chen, C.C. Lin, Water-assisted CO2 laser ablated glass and modified thermal bonding for capillary-driven bio-fluidic application, Biomed. Microdevices 12 (2010) 107–114. [26] C.K. Chung, M.Y. Wu, A hybrid CO2 laser processing for silicon etching, Opt. Express 15 (2007) 7269–7274. [27] G.M. Hale, M.R. Querry, Optical constants of water in the 200-nm to 200-mm wavelength region, Appl. Opt. 12 (1973) 555–563. [28] Marc Madou, in: Fundamentals of Microfabrication, second ed., CRC press, New York, 2002, p. 590. [29] /http://www.engineeringtoolbox.com/overall-heat-transfer-coefficient-d_434. htmlS.