In a multi optical probe confocal imaging system utilizing a microlens arrays as an objective lens, a high numerical aperture is required to improve resolving power. Glass microlens arrays are suitable for high-resolution imaging since they provide outstanding optical properties with a high refractive index. We demonstrated the rapid fabrication of microlens arrays on a high refractive index optical glass substrate via laser assisted thermal imprinting. The optical performance of the fabricated glass microlens arrays were evaluated and compared to that of a polymer microlens. In contrast to the polymer, the real image afforded by, and the calculated resolution of, the imprinted glass microlens arrays were significantly better, at about 0.73 µm compared to the polymer (∼1.56 µm). Our results reveal the considerable potential of direct thermal imprinting as a rapid, single-step, low cost fabrication method for replication of glass microlens array of high dimensional accuracy affording excellent optical performance.
© 2019 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
27 June 2019: A correction was made to the funding section.
Emerging demand for micro-optical devices for numerous optical applications has promoted active research into more efficient and economic ways of micro/nano patterning. The resulting trend is clear, i.e., a shift toward more compact devices with micro/nano structures, high dimensional accuracy and improved performance. Microlens arrays (MLAs) have been widely used for various integrated optical devices, such as sensors  and imaging systems , polarizers , light-emitting diodes (LEDs) and organic light-emitting diodes (OLEDs) [4,5], microfluidic systems , and laser beam homogenizers . In particular, these MLAs are crucial elements in high-imaging throughput with large field of view (FOV) such as used in confocal microscopy imaging system . In addition, MLAs can be integrated to CMOS single photon avalanche diode detector arrays to improve the spatial uniformity of detection efficiency [8,9]. High-imaging throughput with large FOV was possible since the optical information can be obtained simultaneously by each microlens acting as objective lens within the array.
In a multi optical probe confocal imaging system using a MLAs as an objective lens, a high numerical aperture (NA) is required to improve resolving power. Figure 1(a) shows schematic diagrams of the multi-optical probe confocal imaging system using the objective MLAs.
The aperture in objective-side telecentric relay lens, acting as a pinhole for the confocal imaging system, blocks optical information from out of the focal point as shown in Fig. 1(b).
Within the same geometry, glass objective MLAs can achieve higher resolution due to high refractive index compare to polymer MLAs from calculated resolution as shown in Fig. 1(c). Although, glass MLAs can be a suitable candidate for the high-resolution imaging including the confocal imaging, the materials of choice for MLAs has been typically polymers. In the case of polymeric materials, there are various well-established methods to produce MLAs, including injection molding [10,11], ultraviolet (UV) imprinting [12,13], hot embossing [14,15] and roller embossing [16,17]. Even though these polymer-based MLAs are generally superior to glass in terms of production cost and ease of fabrication, they are restricted with respect to thermal and mechanical stability, absorb moisture easily, and have relatively low refractive indices, optical transmissions, and wear resistances . The severity of such optical durability issues increases when these devices are used in applications to diode lasers for beam-shaping or homogenization, which generally require high power . As polymer optics have low laser damage thresholds (LDTs), irreversible damage, commonly known as laser-induced damage (LID) in polymeric optical components, occurs when laser irradiation reaches a sufficient level .
Among the possible candidates for coping with such harsh conditions, glass materials provide outstanding optical properties compared to polymer materials. Furthermore, glass materials can provide superior imaging performance. Compared to polymers, a wide variety of glass types are available; those with a high refractive index feature high-level NA; these glass types are thus candidates for inclusion in high-resolution microscopes. Several methods have been developed to fabricate glass MLAs. State-of-the-art fabrication using standard semiconductor processing technology, such as photolithography and subsequent chemical etching, is only affordable in the case of imprint or master mold fabrication due to its low throughput and requirement for specialized equipment [21,22]. Direct laser structuring had been proposed but the process is slow, and it is difficult to obtain intricate patterns without the use of post-process treatments such as wet etching [23–27]. Machining methods such as diamond cutting have also been reported but are time consuming and offer limited pattern resolution . Most glass types used in the cited studies had a relatively low refractive index (about 1.5); these included fused silica, soda lime, and BK7 glass. Fabrication of glass MLAs using a hybrid sol-gel process has also been reported [29–31]. However, the preparation steps prior to imprinting (precursor solution synthesis and spin-coating) are time-consuming. In addition, light may be lost via reflection from the hybrid material.
To date, direct patterning processes such as hot embossing and compression molding on glass substrates have been recognized as reliable one step-patterning processes for replication of micro/nano structures in a wide range of glass-based materials. However, the total cycle time in most reports ranges from several minutes to hours, depending on the glass substrate used. Conventional compression molding delivers isothermal heat to both the mold and glass, up to and above the glass transition temperature, using external heat sources such as infrared devices. In a recent report on fabrication of glass MLAs, the total cycle time was more than 20 min, including heating, soaking, pressing, and cooling [32–36].
In this paper, we proposed a one-step fine patterning of glass MLAs using direct thermal imprinting assisted by CO2 laser irradiation. Figure 2(a) shows the schematic of the laser-assisted glass imprinting process. In the proposed method, the glass was first externally preheated to slightly below its glass transition temperature at the heating station, and then the temperature throughout the shallow depth of the glass surface was temporarily raised by surface energy absorption through CO2 laser irradiation. The glass sample was then immediately transferred to the press station and subsequently imprinted by metallic mold for a very short contact imprinting time, of about 3 s. Figure 2(b) shows the increase in the surface temperature of the K-PSFn214 optical glass measured by the infrared pyrometer after CO2 laser irradiation at a laser power of 30 W in continuous waveform (CW) mode. The irradiated laser beam diameter was 5 mm at 1/e2 and the measuring spot size of the pyrometer was 3.4 mm. The average temperature within the measuring spot was also measured using the pyrometer. To investigate further the temperature uniformity across the entire glass surface, we used an infrared camera (A655sc; FLIR, USA). The laser-irradiated area was rapidly heated to above 750 ℃ and less than 1 mm of radial heat conduction was observed.
The proposed imprinting process is as follows. Firstly, the glass substrate is preheated to a preheating temperature for 3 min. Then, CO2 laser is irradiated to the preheated glass for 3 s to raise the surface temperature of the glass to above Tg.
Subsequently, the glass is immediately transferred to the imprinting system. After imprinting step for 3s, the cooling of the glass to the preheating temperature for 10 s is required. Finally, the glass is slowly cooled down to the room temperature for 3 min. The key aspect of this technique that differentiates it from the conventional method is that we directly localize the heat incident on the optical glass surface with the CO2 laser prior to the imprinting step, and bring both the heating and cooling cycle outside of the imprinting process. This method facilitates the filling of the glass material into the concave MLAs mold cavities due to the laser driven surface heating effects, and reduces the overall cycle time as shown in Fig. 2(b). We characterized the quality of the imprinted high refractive index glass MLAs using a surface profiler, scanning electron microscope (SEM) and atomic force microscope (AFM). The imprinted lenslet array serves as a micro-objective lens and was integrated into a confocal microscope imaging system. Finally, the imaging performance of the glass MLAs was compared to that of a polymer with the same lenslet design using our in-house confocal microscopy imaging setup. We confirmed that the imaging resolution from the high refractive index glass is significantly better than that of the polymer lens. Our work demonstrates that direct thermal imprinting is an efficient fabrication method for producing glass MLAs with a high optical performance.
2. Materials and methods
2.1 Mold fabrication and materials
Nickel molds with concave MLAs patterns were fabricated in-house using a conventional photolithography process and thermal reflow on the silicon master, followed by nickel seed layer evaporation and electroforming. The MLAs was designed and fabricated with diameter, pitch and sag height of 80, 100 and 15.2 µm, respectively. First, a positive photoresist (AZ GXR 650; AZ Electronic Materials, USA) was spin-coated onto a silicon substrate and soft baked at 105 °C for 70 s. I-line UV light was then exposed at 900 mJ/cm2 using a quartz mask and mask aligner (MA-6; SUSS MicroTec, Germany). The substrate was then immersed in a positive developer (AZ 300 MIF Developer, AZ Electronic Materials) for 3 min. The fabricated micro-cylindrical pedestal was reflowed on a hotplate at a temperature of 160 °C for 30 s to fabricate a spherical MLAs master. A nickel seed layer with a thickness of 50 nm was then deposited onto the prepared silicon master by electron beam evaporation, followed by nano-electrodeposition. The thickness of the electroformed nickel mold was 600 µm, with a patterned area composed of 30 mm × 30 mm. To prevent the adhesion of glass to the mold during thermal imprinting and demolding, 0.5 µm of diamond-like carbon (DLC) was coated onto the nickel stamp of the fabricated MLAs. The optical glass material used in the experiment is K-PSFn214 (Sumita Inc., Japan); the Tg is 425°C, and the refractive index 2.144 at a wavelength of 587.6 nm. To compare imaging performance, we prepared a polymer MLAs made of a UV-curable urethane acrylate-based photopolymer (U088; SK Chemicals Co., Ltd., Korea) on glass slides; the refractive index was 1.522 at a wavelength of 587.6 nm. We replicated polymer MLAs by UV imprinting using the same nickel mold as used for the glass MLAs.
2.2 Laser-assisted glass imprint setup and process sequence
Briefly, this custom-made facility consists of heating, imprinting and cooling stations. First, at the heating station, the optical glass samples, each with dimensions of approximately 10 mm × 10 mm × 1.1 mm, were located on the lower heater block and preheated to a temperature below the Tg of the glass. The optical glass samples were heated by conductance from the bottom through the lower heating block located on a moving linear moving stage. Then, the 2.5 mm diameter beam from the CO2 laser source (V30i, 30 W; Synrad, USA) was expanded using a variable magnification beam expander (BXZ-10.6-2-8X, Ronar Smith, Singapore) and irradiated the top surface of the optical glass. The 10.6 µm wavelength of the CO2 laser was highly absorbed by the glass surface, which promoted substantial thin heating at the glass surface. To control and monitor the laser driven heating process, we used an external computer-controlled program to set the desired laser beam intensity profile. A specific infrared pyrometer for the glass surface temperature measurement (OPTCTG5L10; Optris, Germany) with a spectral response of 5.14 µm was mounted at a distance of 150 mm from the glass surface so that we could monitor the increase in temperature of the glass surface. After predetermining the laser irradiation intensity and duration, we quickly transferred the lower heater block, which was mounted on a moving stage, to the embossing station. Then, a pressurization system monitored by a load cell was immediately lowered onto the upper mold so that we could transfer the pattern onto the optical glass substrate. The embossing pressure and imprinting contact time were 1.5 MPa and 3 s, respectively.
2.3 Confocal imaging set-up with imprinted glass MLAs
We evaluated the optical imaging performance of imprinted MLAs samples using a custom-built confocal microscopy imaging setup, as shown in Fig. 1(a). The optical performance of the imprinted high refractive index K-PSFn214 glass was evaluated by using the MLAs as a micro objective lens array for confocal microscopy imaging. The MLAs sample was illuminated using a white LED light source with an output power of 90 W with peak wavelength of 587.6 nm. The FOV of the confocal imaging system was 1 mm × 1 mm and the resolution was 1 µm / pixel. Sample scanning was performed using a nano-positioning stage. The objective-side relay lens projected an image of each MLAs onto a charge-coupled device (CCD) camera. The measured array area was a 10 × 10 with the scanning range of 100 µm. The advantage of the proposed imaging system is that the aperture of the telecentric relay lens is used as the pinhole of the confocal system. Thus, it is possible to align it by placing the MLA on the object plane of the telecentric relay lens.
3. Results and discussion
3.1 The temperature response of glass by the intensity of CO2 laser irradiation
A thin layer of the glass surface was heated by well-controlled CO2 laser irradiation prior to the embossing step to enhance the material flow into the MLA concave mold cavities. The thickness of the thin layer, defined as the deformation layer, should exceed the required pattern depth; it should be thicker than the sag height of the MLAs in our experiment. Here, the deformation layer is the region in which the temperature exceeds Tg during imprinting. In conventional thermal imprinting, the entire bulk of the glass is heated to a temperature greater than Tg of the glass; hence, the deformation layer occupies the thickness of the glass sample. However, when using the proposed laser-assisted thermal imprinting, only the surface of the glass samples increased beyond Tg. To avoid glass fracture due to the sudden temperature increase induced by the laser source, the bulk of the optical glass was first preheated to slightly below Tg. After several preliminary trials, the initial preheating temperature of the K-PSFn214 optical glass was set to 400°C. Most glass materials inhibit strong optical absorption to irradiation with wavelengths of 10.6 µm; thus, substantial thin heating on the glass surface can be initiated without affecting the bulk of the glass. This temporary effect of heat energy absorption increases the temperature of the glass surface and is also transferred through the glass by heat conduction, causing it to soften and reducing its viscosity, primarily at the surface. Two crucial parameters for the selection of the laser beam intensity profile are the laser power and the irradiation time. In this study, it was most important to increase the glass surface temperature sufficiently for the subsequent embossing procedure without damaging the glass. We identified two possible approaches, which we loosely categorized as high and moderate laser intensity profiles. At high laser irradiation intensities, even though the glass maximum surface temperature increase could be accelerated by the laser irradiation, it also cooled quickly after the laser was switched off. In other cases, we observed bulk deformation on the glass when the laser irradiation continued for too long. In contrast, a medium laser intensity is preferable because the glass surface can be heated more slowly, thus slowing the cooling rate and allowing the glass surface to maintain its low viscosity for a longer time, until the glass is transferred to the imprinting station for replication.
Figure 3(a) shows the increase in the glass surface temperature with respect to the laser irradiation time using a laser power of 30 W and a beam diameter of 5 mm. Based on the temperature history profiles recorded by the infrared pyrometer, the optimal processing condition for high replication quality was determined. Overall, the viscosity of the glass surface decreases as the temperature of the glass surface increases, thus improving the filling ratio of the glass onto the concave MLA cavities. After a laser irradiation time of 1 s, the glass surface temperature at the start of imprinting was approximately 480°C. No replication occurred under this condition, as shown in Fig. 3(b). We believe that the temperature dropped to just Tg or below due to the additional heat loss at the glass surface during the transfer process. In the case of 2 s of laser irradiation time, the MLA replication height reached 84.5%. For 3s, full cavity filling was achieved. A similar result was obtained when a laser irradiation time of 4 s was used. The maximum increase in the glass surface temperature appeared to become saturated at laser irradiation times of 5 s and 6 s. However, we clearly observed bulk deformation of the glass sample when the laser irradiation time increased beyond 7 s. Based on these results, we considered a laser irradiation time of between 3 s and 4 s prior to the imprinting step to be optimal. However, when the laser irradiation was not sufficient, the filling ratio was incomplete, even under higher imprinting pressures, such as 4 MPa. We also observed that longer imprinting contact times, greater than 3 s, are not necessary because the temperature dropped to the initial preheating temperature, after which no further material flow occurred.
In the case of K-PSFn214 glass, the surface temperature attained was 790 ± 10 °C after 3 s of laser irradiation time, with an average heating rate of 130 °C/s. Considering the transfer time of 1s, a temperature of approximately 650 °C at the start of the glass contact embossing was sufficient to decrease the viscosity of the surface, thus improving the filling of the glass material into the microstructured cavities. Once the mold was in contact with the glass material under load during the embossing step, a very fast temperature drop, to the initial temperature of the glass surface, was expected due to the heat conductance at the interface. Therefore, in this method, demolding could be performed after a very short contact embossing time.
3.2 Imprinted glass MLAs pattern fidelity
Figure 4(a) shows the SEM image of the glass sample imprinted using our new method. MLAs with diameter of 80 µm, pitch of 100 µm, and sag height of 14.8 µm were faithfully replicated on K-PSFn214 optical glass under laser power of 30 W delivered for 3 s; prior to imprinting, the beam diameter was 5 mm at 1/e2. The imprinting pressure and contact time were approximately 1 MPa and 3 s, respectively. The SEM image confirm that the glass MLAs was successfully imprinted and demolded without any visible damage to the pattern; the DLC coating served as an effective anti-adhesion layer in the glass/mold interface.
Lens sag height and the profile of the imprinted MLAs were measured using a stylus profilometer (Dektak XT; Vecco Instruments, USA). The profiles of the mold and imprinted glass matched very closely, indicating high-fidelity replication as shown in Fig. 4(b).
The profile of the imprinted glass MLAs was generally slightly smaller than that of the mold, as would be expected given glass shrinkage after cooling; the coefficients of thermal expansion of the mold and glass differ. However, the deviation (approximately 0.4 µm) was very small compared to the nominal MLAs geometry as shown in Fig. 4(c). Glass MLA sag heights across 5 mm diameter of imprinted area was uniform; this was confirmed by measuring imprinted sag heights randomly at five different points on the sample as listed in Table 1. We used AFM to evaluate the surface morphology of the top and center of the imprinted glass MLAs; the measurement area was 1 µm × 1 µm. The average roughness (Ra) and root mean square (RMS) roughness of the imprinted glass were approximately 1 and 3 nm, respectively, indicating fine optical finishes.
Transmittance of the optical glass before and after thermal imprinting was measured using an ultraviolet-visible-near infrared (UV-Vis-NIR) spectrophotometer. Raw K-PSFn214 optical glass exhibited a transmittance > 95% at wavelengths of 500–2,000 nm. After thermal imprinting, no significant change in transmittance was evident. Thus, imprinting did not affect structural stability. The transmittance of many other types of optical polymer is generally only about 85%, and several absorption peaks within similar wavelength ranges are seen . The optical transmittance of glass directly imprinted to form MLAs was better over a much broader wavelength than that of the polymer, enhancing optical resolution and thus improving imaging.
3.3 Confocal microscopy imaging of MLAs made of highly refractive glass
To obtain proof-of-concept, the imaging performance of glass MLAs was evaluated via confocal microscopy as described in section 2.3. First, we used basic optical theory equations to estimate the radius of curvature, focal length, and NA of the microlens. The radius of curvature of a spherical microlens is:
The lateral resolution of a single microlens can be calculated as follows:
We used the USAF-1951 Resolution Test Target (Edmund Optics, USA) to further confirm the superior confocal imaging afforded by glass compared to polymer MLAs as shown in Fig. 6(a) and 6(b), respectively. From an optical simulation of our confocal imaging system, we confirmed that the on-axis point information of the individual microlenses was focused on the image sensor, without cross-talk. Furthermore, we experimentally confirmed the same results. The glass MLAs distinguished the minimum linewidth of 1.55 µm (element 3 of group 8) but the polymer MLAs did not. Figure 6(c) quantitatively compares intensity profiles along the dashed lines of the five-fold magnified inset images in Fig. 6(a) and 6(b). The modulation transfer function (MTF) of the glass MLAs was much higher than that of the polymer MLA.
To explore imaging performance further, we imaged an OLEDs thin-film transistor (TFT) glass with a microscale pattern. Figures 7 shows the image obtained using the highly refractive index glass MLAs; the inset images are five-fold magnifications of the image.
The image obtained using the glass MLAs is of higher resolution and contrast. The edge and boundary of the OLEDs TFT glass were clearly visible but uniformly blurry when the polymer MLA operated across a similar field of view. It is thus clear that highly refractive index glass MLAs prepared using our method can be used for high-resolution imaging, where both durability and mechanical resistance are very important.
For the current MLAs imprinted on highly refractive index glass, the calculated NA is high (approximately 0.51). The geometric MLAs could be tailored to achieve a higher NA, for example by increasing the sag height and decreasing the radius of curvature. The largest possible acceptance angle is that achievable when the ratio of the lens diameter to the lens sag height, D/h, is 2; it is technically challenging to achieve this ratio using existing microfabrication technologies. In other words, to improve the imaging performance of the polymer MLAs, more precise design and fabrication strategies are required.
Although we do not address the topic here, the use of single-wavelength light sources such as lasers will improve resolution and simultaneously increase the damage threshold compare to the polymer. Our MLAs can also be used when high NA values are required, such as during endoscopic imaging.
We imprinted a high-quality MLAs onto high refractive index optical glass via rapid thermal imprinting with a very short contact imprinting time of approximately 3 s, having applied CO2 laser irradiation prior to the imprinting step. We investigated the quality of the imprinted pattern and observed a very smooth surface finish with a close match to the profile of the mold. The fabricated lenslet arrays were then integrated into a confocal microscopy imaging system, in which they acted as micro objective lens arrays. This enabled us to evaluate their imaging performance. The experimental and theoretical lateral resolutions at the FWHM of the glass MLAs for a FOV of 1 mm × 1 mm were in good agreement, at 0.74 and 0.70 µm, respectively. We also evaluated the resolving power of the confocal imaging system using the 1951 USAF resolution test target. As expected, the high refractive index glass achieved a higher resolution than the polymer MLAs. In images of reconstructed OLEDs TFT obtained using the glass MLAs, the edges and boundaries were clearly visible. We believe that our results confirm the utility of low-cost, high-throughput direct thermal imprinting for fabrication of high-quality optical components.
National Research Foundation of Korea (NRF) (2015R1A5A1037668), Ministry of Trade, Industry and Energy (MOTIE) (Grant N0002310).
1. R. Kasztelanic, A. Filipkowski, D. Pysz, R. Stepien, A. J. Waddie, M. R. Taghizadeh, and R. Buczynski, “High resolution Shack-Hartmann sensor based on array of nanostructured GRIN lenses,” Opt. Express 25(3), 1680–1691 (2017). [CrossRef]
2. W. Choi, R. Shin, J. Lim, and S. Kang, “Design methodology for a confocal imaging system using an objective microlens array with an increased working distance,” Sci. Rep. 6(1), 33278 (2016). [CrossRef]
3. Y. Lee, C. Yu, and J. Kim, “Polarizer-free liquid crystal display with double microlens array layers and polarization-controlling liquid crystal layer,” Opt. Express 23(21), 27627–27632 (2015). [CrossRef]
4. C. Liu and G. J. Su, “Enhanced light extraction from UV LEDs using spin-on glass microlenses,” J. Micromech. Microeng. 26(5), 055003 (2016). [CrossRef]
5. F. Galeotti, W. Mróz, G. Scavia, and C. Botta, “Microlens arrays for light extraction enhancement in organic light-emitting diodes: A facile approach,” Org. Electron. 14(1), 212–218 (2013). [CrossRef]
6. K. W. Ro, K. Lim, B. C. Shim, and J. H. Hahn, “Integrated Light Collimating System for Extended Optical-Path-Length Absorbance Detection in Microchip-Based Capillary Electrophoresis,” Anal. Chem. 77(16), 5160–5166 (2005). [CrossRef]
7. I. Sohn, H. Choi, Y. Noh, J. Kim, and M. S. Ahsan, “Laser assisted fabrication of micro-lens array and characterization of their beam shaping property,” Appl. Surf. Sci. 479(15), 375–385 (2019). [CrossRef]
8. G. Intermite, A. McCarthy, R. E. Warburton, X. Ren, F. Villa, R. Lussana, A. J. Waddie, M. R. Taghizadeh, A. Tosi, F. Zappa, and G. S. Buller, “Fill-factor improvement of Si CMOS single-photon avalanche diode detector arrays by integration of diffractive microlens arrays,” Opt. Express 23(26), 33777–33791 (2015). [CrossRef]
9. J. M. Pavia, M. Wolf, and E. Charbon, “Measurement and modeling of microlenses fabricated on single-photon avalanche diode arrays for fill factor recovery,” Opt. Express 22(4), 4202–4213 (2014). [CrossRef]
10. K. Pang, L. Song, F. Fang, Y. Zhang, and H. Zhang, “An imaging system with a large depth of field based on an overlapped micro-lens array,” CIRP Ann. 65(1), 471–474 (2016). [CrossRef]
11. M. Chakrabarti, C. Dam-Hansen, J. Stubager, T. F. Pedersen, and H. C. Pedersen, “Replication of optical microlens array using photoresist coated molds,” Opt. Express 24(9), 9528–9540 (2016). [CrossRef]
12. K. Kawahara, T Kikuchi, S. Natsui, and R. O. Suzuki, “Fabrication of ordered submicrometer-scale convex lens array via nanoimprint lithography using an anodized aluminum mold,” Microelectron. Eng. 185-186(5), 61–68 (2018). [CrossRef]
13. J. Chen, J. Cheng, D. Zhang, and S. Chen, “Precision UV imprinting system for parallel fabrication of large-area micro-lens arrays on non-planar surfaces,” Precis. Eng. 44, 70–74 (2016). [CrossRef]
14. H. Xiong and Z. Wang, “Fabrication of Chalcogenide Microlens Array Using Hot Embossing Method,” in 2018 IEEE Sens. 1–3 (2018).
15. J. Chen, C. Gu, H. Lin, and S. Chen, “Soft mold-based hot embossing process for precision imprinting of optical components on non-planar surfaces,” Opt. Express 23(16), 20977–20985 (2015). [CrossRef]
16. C. Chang and M. Tsai, “Development of a continuous roll-to-roll processing system for mass production of plastic optical film,” J. Micromech. Microeng. 25(12), 125014 (2015). [CrossRef]
17. C. Chang and J. Chu, “Innovative design of reel-to-reel hot embossing system for production of plastic microlens array films,” Int. J. Adv. Des. Manuf. Technol. 89(5-8), 2411–2420 (2017). [CrossRef]
18. H. Ottevaere, R. Cox, H. P. Herzig, T. Miyashita, K. Naessens, M. Taghizadeh, R. Völkel, H. J. Woo, and H. Thienpont, “Comparing glass and plastic refractive microlenses fabricated with different technologies,” J. Opt. A: Pure Appl. Opt. 8(7), S407–S429 (2006). [CrossRef]
19. O. Homburg, A. Bayer, T. Mitra, J. Meinschien, and L. Aschke, “Beam shaping of high power diode lasers benefits from asymmetrical refractive micro-lens arrays,” International Society for Optics and Photonics In High-Power Diode Laser Technology and Applications VI (6876), 68760B International Society for Optics and Photonics (2008).
20. K. Kiedrowski, J. Thiem, F. Jakobs, J. Kielhorn, I. Balasa, D. Kracht, and D. Ristau, “Determination of the laser-induced damage threshold of polymer optical fibers,” In Laser-Induced Damage in Optical Materials 2018: 50th Anniversary Conference, 10805, 108052C International Society for Optics and Photonics (2018).
21. P. Zhang, G. Londe, J. Sung, E. Johnson, M. Lee, and H. J. Cho, “Microlens fabrication using an etched glass master,” Microsyst. Technol. 13(3-4), 339–342 (2006). [CrossRef]
22. X. Huang, P. Wang, E. Lin, J. Jiao, X. Wang, Y. Li, Y. Hou, and Q. Zhao, “Fabrication of the glass microlens arrays and the collimating property on nanolaser,” Appl. Phys. A: Mater. Sci. Process. 122(7), 649 (2016). [CrossRef]
23. F. Chen, Z. Deng, Q. Yang, H. Bian, G. Du, J. Si, and X. Hou, “Rapid fabrication of a large-area close-packed quasi-periodic microlens array on BK7 glass,” Opt. Lett. 39(3), 606–609 (2014). [CrossRef]
24. Y. Wei, Q. Yang, H. Bian, F. Chen, M. Li, Y. Dai, and X. Hou, “Fabrication of high integrated microlens arrays on a glass substrate for 3D micro-optical systems,” Appl. Surf. Sci. 457, 1202–1207 (2018). [CrossRef]
25. S. Tong, H. Bian, Q. Yang, F. Chen, Z. Deng, J. Si, and X. Hou, “Large-scale high quality glass microlens arrays fabricated by laser enhanced wet etching,” Opt. Express 22(23), 29283–29291 (2014). [CrossRef]
26. D. Nieto, M. T. Flores-Arias, G. M. O’Connor, and C. Gomez-Reino, “Laser direct-write technique for fabricating microlens arrays on soda-lime glass with a Nd: YVO 4 laser,” Appl. Opt. 49(26), 4979–4983 (2010). [CrossRef]
27. F. Chen, H. Liu, Q. Yang, X. Wang, C. Hou, H. Bian, W. Liang, J. Si, and X. Hou, “Maskless fabrication of concave microlens arrays on silica glasses by a femtosecond-laser-enhanced local wet etching method,” Opt. Express 18(19), 20334–20343 (2010). [CrossRef]
28. P. Li, J. Xie, J. Cheng, and Y. N. Jiang, “Study on weak-light photovoltaic characteristics of solar cell with a microgroove lens array on glass substrate,” Opt. Express 23(7), A192–A203 (2015). [CrossRef]
29. S. D. Zilio, G. D. Giustina, G. Brusatin, and M. Tormen, “Microlens arrays on large area UV transparent hybrid sol–gel materials for optical tools,” Microelectron. Eng. 87(5-8), 1143–1146 (2010). [CrossRef]
30. X. C. Yuan, W. X. Yu, M. He, J. Bu, W. C. Cheong, H. B. Niu, and X. Peng, “Soft-lithography-enabled fabrication of large numerical aperture refractive microlens array in hybrid SiO2–TiO2 sol-gel glass,” Appl. Phys. Lett. 86(11), 114102 (2005). [CrossRef]
31. M. He, X.-C. Yuan, N. Q. Ngo, J. Bu, and V. Kudryashov, “Simple reflow technique for fabrication of a microlens array in solgel glass,” Opt. Lett. 28(9), 731–733 (2003). [CrossRef]
32. Y. K. Kim, J. H. Ju, and S. M. Kim, “Replication of a glass microlens array using a vitreous carbon mold,” Opt. Express 26(12), 14936–14944 (2018). [CrossRef]
33. C. Huang, W. Hsiao, K. Huang, K. Chang, H. Chou, and C. Chou, “Fabrication of a double-sided micro-lens array by a glass molding technique,” J. Micromech. Microeng. 21(8), 085020 (2011). [CrossRef]
34. Y. Chen, Y. Y. Allen, D. Yao, F. Klocke, and G. Pongs, “A reflow process for glass microlens array fabrication by use of precision compression molding,” J. Micromech. Microeng. 18(5), 055022 (2008). [CrossRef]
35. G. C. Firestone and A. Y. Yi, “Precision compression molding of glass microlenses and microlens arrays—an experimental study,” Appl. Opt. 44(29), 6115–6122 (2005). [CrossRef]
36. H. Mekaru, T. Tsuchida, J. Uegaki, M. Yasui, M. Yamashita, and M. Takahashi, “Micro lens imprinted on Pyrex glass by using amorphous Ni–P alloy mold,” Microelectron. Eng. 85(5-6), 873–876 (2008). [CrossRef]
37. N. G. Sultanova, S. N. Kasarova, and I. D. Nikolov, “Characterization of optical properties of optical polymers,” Opt. Quantum Electron. 45(3), 221–232 (2013). [CrossRef]