Chapter 3 Design of Lenticular-lens-based Micro-optical Structure for
3.6 Simulated Results and Discussions
3.6.2 Double-screen Function
(a) (b)
(c) (d)
Fig. 3-9. Angular distribution of proposed 3D function for (a) left and (b) right eye’s image, and (c) and (d) are their respected cross-section intensity distribution in horizontal orientation.
3.6.2 Double-screen Function
By using ASAP to simulate the design model, the optimized angular distribution of double-screen function has been found for bottom and top screens as shown in Fig.
3-10. The valid viewing angle is derived ranging from approximately 0o to 40o for each
there are still a large amounts of light transmitted with more than 50% of efficiency as shown in red center region of Figs. 3-9 (a) and (b). The top and bottom screens have the similar light efficiency as shown in Figs. 3-9 (c) and (d). All in all, since both screens can be functioned at similar viewing angle, the observer could view two images at the same time with 30o viewing freedom vertically.
(a) (b)
(c) (d)
Fig. 3-10. Angular distribution of proposed double-screen function for (a) bottom and (b) top direction image, and (c) and (d) are their respected cross-section intensity distribution in vertical orientation.
3.6.3 3D Double-screen Function
The simulated results of 3D double-screen display represent the combination of 3D and double-screen features. Due to human’s eyes perception, the crosstalk needs to be of less than 10 %, hence, viewing-angle ranges are around - 35∘~ 35∘and ± 5∘~
± 40∘in the horizontal direction and vertical direction, respectively, as shown in Fig.
3-11. The angular distribution of light intensity for each screen is shown in Fig. 3-12.
Because the 3D function layer is designed to place on the top of double-screen function layer, the thickness is thicker than operating 3D images alone. Since the 3D function is favorable for bulky substrate, the larger distance between the 3D function lenses arrays and color filter enhanced more views of 3D images. Thus, in the proposed 3D double-screen design, there are roughly 6 views generated. Due to the low light intensity at very left and very right viewing zones, the images at these two zones may not be perceivable. The overall amounts may end up with 4 clear viewing zones horizontally for both top and bottom screens.
Fig. 3-11. Schematic diagram of acceptable viewing-angles for top and bottom screens.
Bottom Screen:
(a) (b)
(c) (d)
Fig. 3-12. Angular distributions of the proposed 3D double-screen display for (a) left and (b) right eye’s images of top screen; (c) left and (d) right eye’s images of bottom screen.
3.7 Summary
The intention of research is to design a lenticular-lenses-based micro-optical structure generating dual 3D images simultaneously. The principles of producing 3D and double-screen images have been investigated. Hence, 3D and double-screen functions have designed to align with color filter in portrait and landscape orientation and to direct the incident light to the desired regions. In addition, the investigation on proposed structure shows 3D function is favoring thicker substrate, smaller material refraction index, smaller pitch and larger lens radius, which is opposite to double-screen function’s favors. After considering the parameters of potential threats
40∘
to the range of viewing angles, the performances of 3D and double-screen functions have been optimized. Moreover, the simulated results of 3D and double-screen functions and their combination have verified the proposed concept. The micro-optical structure consists 3D function layer and double-screen layer are generated at least 4 viewing zones of the clear 3D images with viewing angle range around - 35∘~ 35∘and ± 5∘~ ± 40∘in the horizontal direction and vertical direction, respectively. Thus, by including the planar mirror in the optical system, observers can view 3D images at top and bottom screens simultaneously.
Chapter 4
Fabrication of the Lenticular-lens-based Micro-optical Structure by Excimer Laser Micromachining System
4.1 Introduction
The fabrication technologies of micro-optics are getting more and more advanced.
Fabricating the more flexible components and system designs, such as smaller and precise structures, becomes possible. A large variety of materials is available, ranging from glass to semiconductor materials and on to plastic. The technologies to fabricate micro-optical elements can be simply classified into two main categories, lithographic and non-lithographic techniques. Lithography is the name for a sequence of processing steps, pattern generation, coating or thin layer deposition, alignment, and exposure, pattern transfer for structuring the surface of planar substrates.
In the 1970s planar lithographic fabrication techniques were adapted from semiconductor process to the fabrication of optical elements, for example, to fabricate special beam splitters and lens arrays. The use of these techniques allows one to generate optical component with dimension in the micrometer range. Besides, in an effort to fabricate some specified elements such as micro-lens, non-lithographic techniques, for example, diamond cutting and micro-jet printing have been investigated.
However, these techniques often do not have small enough critical dimensions or are unable to generate the micro-optical element with inadequate designed surface relief.
Other than that, the fabrication time and cost are also the issues.
In terms of recent fabrication technology, the economical processes with satisfying the requirements on fabricating the lenticular-lenses arrays have to be discovered and developed. Since excimer laser ablation is a rapid and effective way for micromachining a surface relief onto a substrate material, the study will be progressed based on utilizing the excimer laser system to fabricate the lenticular-lens-based micro-optical structure.
4.2 Fabrication Principle
In order to make possible of fabricating lenticular-lenses by the excimer laser micromachining system, the process steps, system structure and micromachining principle are explored.
4.2.1 Fabrication Process
There are several fabrication techniques. The most common fabrication processes of lithography technology needs 6 steps as shown in Fig. 4-1. First the photosensitive materials like photoresist are coated on the substrate. The modulated illumination for desired pattern is applied to expose the photoresist. After exposure, a development step converts the exposed photoresist onto a surface profile. In a further processing step, the surface profile of the photoresist pattern can be transferred into the substrate.
Consequently, the micro-optical elements can be fabricated with above procedures on the substrate.
The intention of this research is to fabricate the lenses by excimer laser
substrates are the only steps similar to the general lithography method. The step of excimer laser micromachining combined several lithography processes includes exposure and etching in a single step. PR coating and development are not necessary to be part of process any more. In addition, the fabricated substrate will be cleaned by Ultrasonic Cleaner with Isopropyl Alcohol. The inspection of the patterns can be evaluated through microscope and Zygo’s interferometer.
Fig. 4-1. Detailed fabrication processes of general lithography technology.
Fig. 4-2. Detailed fabrication processes of excimer laser micromachining technology.
4.2.2 Excimer Laser Micromachining System
The excimer laser micromachining system consists several components as shown in Fig.4-3. The utilized laser is a Kr F excimer laser operating at 248nm. The maximum energy per pulse is typically equal to 0.7 J, the pulse duration is
the program. Because the intensity profile of an excimer laser output is quite non-uniform, beam forming optics and beam homogenizer inside beam delivery system are used to produce a uniform intensity field at the mask plane. This step can creates a highly uniform (± 5% RMS) illumination of 12 × 12mm at the mask plane.
Then the mask plane is imaged on the substrate to ablate the polymer substrate with 4x or 10x demagnification by UV objective of a 0.1 or 0.2 NA respectively.
Fig. 4-3. Schematic diagram of excimer laser micromachining system.
4.2.3 Micromachining Technique
The structures are created by scanning the image of the contour mask across the substrate while operating the laser with a fixed pulse repetition rate. The principle of this method is that only the mask contour defines the locally applied laser pulse quantity. The region with more excimer laser passed through the contour mask will result deeper depth. The detail process is shown in Fig. 4-4. More complex structures can be achieved by subsequent scan step. In this way, lenticular-lenses and
lenticular-lenses arrays are expected to be fabricated by scanning in orthogonal directions with a containing semi-circular shape contour mask as shown in Fig. 4-5.
Excimer laser beam
Projection objective Projection mask
Substrate
Scann
ing direction
Fig. 4-4. Prism structure fabricated by scanning with a contour mask.
4.3 Fabrication Stability and Lens Quality
Several variables were considered for obtaining the quality and stability of fabricating lenticular-lenses arrays by excimer laser micromachining system.
4.3.1 Energy Stability
The excimer laser has been detected continuously for the 4 weeks by energy detector as shown in Fig. 4-6. The detected energy is the amount of output before emerging through the condensed lens. Because the detected energy is the reference for us to see the trend of laser stability, the unit between detected energy and input energy are not important. The relationship between input energy and detected energy is linear as shown in Fig. 4-6.
During the fabrication period, the maximum input energy can be obtained was found to be 240 mJ. The detected energy shows the input energy is similar for each week, except the 2nd week was lower. The lower laser energy is due to low gas pressure and can be avoid by changing the KrF gas. In addition, each shot of excimer laser has some tolerance. The deviation, which is defined as the percentage of variance energy different from average laser shot energy, of laser shot in each week has also been evaluated as shown in Fig. 4-7. While input energy is higher, deviation becomes smaller.
Deviations are in the range of 4 ~ 7 % of detected energy for these 4 weeks. As the result, the energy stability of excimer laser can be relied on with keeping an eye on the level of gas pressure.
Fig. 4-6. Excimer laser energy for 4 weeks continuously.
Fig. 4-7. Deviation of laser shot for each week.
4.3.2 Lens Radius with Laser Energy
Fewer variables mean fewer dependents which have brought more convenient in further investigation. We have attempted to convert laser energy includes its input energy and attenuation to the independent constant. In addition, lens radius does not change a lot when laser energy is about 200 mJ as shown in Fig. 4-8. As set up the
after attenuation is 192 mJ, laser energy output of each shot seems to be stable at this level. Hence, the variables of fabrication considerations have reduced to laser penetration rate and stage feeding rate.
Fig. 4-8. Lens radius with laser energy.
4.3.3 Lens Radius with Laser Penetration Rate
Penetration rate is inversely proportional to the fabricated lens radius as shown in Fig. 4-9. Higher penetration rate will have smaller lens radius, rougher lens surface and thinner the substrate. According to the trend line, it seems the number approaching 18 Hz is preferable to the proposed design. However, when the penetration rate reaches 18 Hz, the substrate becomes very thin for the small lens radius in our contour mask design.
The thinnest point may break through the substrate. Therefore, the range of penetration rate has to be as small as possible, at least should be less than 18 Hz, in the experiment.
Thus, we have to use the other variable, such as stage feeding rate to cover the insufficient of lens curvature.
Fig. 4-9. Lens radius with laser penetration rate.
4.3.4 Lens Radius with Stage Feeding Rate
When stage feeding rate is smaller than 2 mm/min., the radius of fabricated lens is sensitive to the stage feeding rate as shown in Fig. 4-10. In order to obtain the expected radius, it has to complement with laser penetration rate. Due to the target of our lens design, which is between 600 um and 800 um, the stage feeding rate has to be less than 3 mm/min. The lower stage feeding rate helps in improving the surface smooth.
Nevertheless, lower the stage feeding rate increases the fabrication period, for example, the stage feeding rate of 0.5 mm/min. needs double time period of 1 mm/min. This is a trade off in the fabrication consideration. Thus, in this experiment, stage feeding rate is chosen from 1 mm/min to 3 mm/min.
Fig. 4-10. Lens radius with stage feeding rate.
4.4 Fabrication Tolerances
The deviations in the fabrication are almost incapable to avoid totally. The possible errors of fabricating by Excimer laser micromachining system are dual curvatures in single lens and lens gap in between lenses. In order to obtain the desired functions, the effects of inaccurate fabricated samples have to be discovered and simulated.
4.4.1 Dual Curvatures Effect
Dual curvatures effect maybe occurred when the excimer laser is not uniformly distributed in the projecting area. If the left or right portion of projecting area has larger amount of laser energy, the resulted lens radius will be smaller. The schematic of singular curvature and dual curvatures lens structures is shown in Fig. 4-11.
Since one portion of lens is taller than another, some light will be blocked by the taller portion. The simulated results of varies pairs of dual curvatures in single lens are shown in Fig. 4-12. While the deviations of lens radius between left and right
portions are similar, the resulted viewing angles are same as the desired function.
Moreover, until the deviations reached 50 um as shown in Fig. 4-12 (d), some light path may be blocked to result narrower viewing angle. The study shows the uncertainty of left and right portions in single lens has to be small as possible, thus, the desired viewing angles can be obtained.
(a) (b)
Fig. 4-11. Schematics of (a) singular curvature and (b) dual curvatures lens structures.
(a) (b) (c) (d)
Fig. 4-12. Simulated results of half lens with radius of 150 um and another half with radius (a) 200 um, (b) 150 um, (c) 140 um and (d) 100 um.
4.4.2 Lens Gap Effect
In the Excimer laser micromachining system, the lenticular-lenses are fabricated one by one. There may be the gaps in between the lenses as shown in Fig. 4-13. Since
the lens design may need to seek for continued lenses and avoid the gaps in between lenses.
(a) (b)
(c) (d)
Fig. 4-13. Schematics of lens structures with (a) no gap and (b) gaps, and their respected simulated results.
4.5 Experiments
The proposed micro-optical elements were fabricated by using Excimer laser micromachining system with contour mask at Instrument Technology Research Center (ITRC) to fabricate. The laser system used is an Excitech 7000 series excimer laser workstation as shown in Fig. 4-14.
Fig. 4-14. Appearances of Excitech7000.
Energy of excimer laser, number of micromachining pulses so called # of shot, and laser repeat rate can be precisely controlled to our need. We chose the projection system with 10x because the higher demagnification increases the resolution of the microoptical elements fabricated by using contour mask, i.e. the minimum pixel size on the elements shrink. NA is 0.2. The material of work is a PC polymer (polycarbonate) of the thickness 0.5 µm (purchased from Goodfellow).
The micromachining parameters and target sample parameters are shown in Tables 4-1 and 4-2. The outlook of target sample is shown in Fig. 4-15.
After excimer laser micromachining, the radius of fabricated lenses and its angular distribution of light intensity are measured by using Zygo’s optical interferometer as shown in Fig. 4-16 and Conoscopic system as shown in Fig. 4-17.
Table 4-1. Micromachining parameters.
Parameters Values range
Energy 240 mJ/pulse
Attenuation 0.8 Laser penetration rate 10 ~ 18 Hz
Stage feeding rate 1 ~ 3 mm/min.
Table 4-2. Fabrication targets for the lenticular-lenses arrays.
Fig. 4-15. Outlook of a set of lenticular-lenses to be fabricated.
H
X
W
L C
D
Fig. 4-16. Appearances of Zygo’s optical interferometer.
Fig. 4-17. Appearances of ELDIM EZContrast 160 measurement system.
4.6 Experimental Results and Discussions
The experiment is taken based on verifying the simulated results with the reality on lens radius and viewing angle for both functions.
4.6.1 Lens Radius
The sample of lenticular-lenses-based micro-optical structure with the diagonal of 0.54 inch, as shown in Fig. 4-18, is mainly fabricated by Excimer laser micromachining system. The lenticular-lens pitch and length are 3.6 mm and 11 mm, respectively. By utilizing Zygo’s interferometer to inspect the fabricated substrate, the results of lens thick and lens half width, as shown in Fig. 4-19, can be used with equation 3-7 to calculate the lens radius as shown in Table 4-3.
From the experimental results, we can find that trial 3 and trail 13 are close to the design for double-screen and 3D functions, respectively. In addition, by using the similar settings as trails 2 and 4, excimer laser has fabricated the lenses twice with similar output results within + 50 um. The smooth of surface is less than 1 um which is acceptable for our design. In the other words, the excimer laser has high possibility of reproducing the lenses. Comparing to double-screen function, 3D function layer is more difficult and needs more time to fabricate due to its smaller pitches dimensions.
Nevertheless, 3D and double-screen function layers can be fabricated by excimer laser with the laser energy equals 240 mJ, attenuation 0.8 and stage feeding rate 1 mm/min.
The only variable is penetration rate which for 3D and double-screen function layers are 10 Hz and 15 Hz, respectively.
Fig. 4-18. Photograph of sample fabricated lenticular-lens for double-screen function.
11mm
3.6 mm
Table 4-3. Fabrication parameters and resultant lenses radii of each trail.
4.6.2 Viewing Angle for 3D and Double-screen Functions
The simulation for the top screen of the double-screen function yields the viewing angles of - 45 ∘ ~ 45 ∘ and - 5 ∘ ~ 40 ∘ in horizontal and vertical directions, respectively. After testing the sample by Conoscopic System, the measured viewing angles are - 60∘~ 60∘and - 10∘~ 40∘in the horizontal and vertical directions, respectively. The simulated and experimental results for double-screen function are compared in Fig. 4-20. The experimental results show the wider viewing angles in the horizontal direction. Since double-screen function used to direct the light paths in vertical directions, the wider of horizontal direction does not affect the results.
The experimental result of 3D function is wider than the simulated result, but they have the similar feature, 3 viewing screens. There are several possible reasons of having different magnitudes with the similar trend. The obviously reason is that the lambertian light’s incident angle of Conoscopic System is larger than the settings of the simulation. Moreover, the coordinate axes of the simulation software and measuring
system are different. Simulation software has coordinate axes in X and Y. The Conoscopic System has the axes in Phi and Theta. Since the Phi and Theta can project on X and Y axes, the relationship between both coordinates’ measured results will have similar trend. The results show similar tendency of using different coordinate systems, but the magnitude of the results will not be the same values. Furthermore, the other reason may be the difficulty of alignment for the small dimensional objects.
Due to the necessary of blocking the opponent directional light, the barrier has to locate at precise position and only let the designated directional light passes through.
Due to the tiny aligning point and colorless substrate, there are some crosstalk happened.
4.7 Summary
A prototype of micro-optical structure for 3D and double-screen functions was fabricated by the excimer laser micromachining system. The fabrication parameters, such as laser energy, attenuation, penetration rate and stage feeding rate are found to be 240 mJ/pulse, 0.8, 10 Hz ~18 Hz and 1 mm/min. ~ 2 mm/min., respectively. As the result, the lens radius can be estimated from the measured half lens pitch and lens thickness. The radii of fabricated lenticular-lenses for 3D and double-screen functions are within the tolerances of design values, 800 + 50 um and 650 + 50 um, respectively.
A prototype of micro-optical structure for 3D and double-screen functions was fabricated by the excimer laser micromachining system. The fabrication parameters, such as laser energy, attenuation, penetration rate and stage feeding rate are found to be 240 mJ/pulse, 0.8, 10 Hz ~18 Hz and 1 mm/min. ~ 2 mm/min., respectively. As the result, the lens radius can be estimated from the measured half lens pitch and lens thickness. The radii of fabricated lenticular-lenses for 3D and double-screen functions are within the tolerances of design values, 800 + 50 um and 650 + 50 um, respectively.