1. Introduction
Liquid crystals (LCs) have been widely used in recent decades owing to their giant optical nonlinearity, which is nine orders of magnitude larger than ordinary materials [
1,
2,
3]. This optical nonlinearity is attributed to the orientational change in LC molecules when irradiated with a polarized light source at a low light intensity of 100 W/cm
2, which is relatively lower than the materials used in the past. Therefore, this photoinduced reorientation of LC molecules upon irradiation with a light source is classified as a nonlinear optical (NLO) effect that has recently attracted much attention owing to their high sensitivity and significant change in refractive index [
4,
5,
6]. Many scientists have been intrigued by finding different innovative approaches to enhance the optical nonlinearity of LC materials and observed that this can be accomplished by doping specific dye molecules in the LC materials [
7,
8]. Jánossy et al. first reported that an anthraquinone dye could serve as a photo-functional dye to significantly enhance the sensitivity of molecular reorientation [
9,
10,
11]. Since then, many more approaches have been established and proposed to improve this optical nonlinearity of dye-doped LC molecules, for instance, by introducing polymer-stabilized LC (PSLC) systems [
12,
13].
Previous research reports have shown that polymer stabilization can enhance the thermal and electro-optical properties of LC molecules [
14,
15]. Aihara et al. reported that polymer stabilization improved the orientational optical nonlinearity by 56%; that is, an improvement of six orders of magnitude of threshold intensity to cause a photoinduced molecular reorientation [
12]. The photoinduced molecular alignment of PSLC systems was further enhanced by manufacturing hybrid systems containing homogeneously aligned LC molecules on one side and homeotropically aligned LC molecules on another side [
16], increasing the polymer concentration [
13], optimizing the silane coupling concentration that induces the homeotropic alignment of LC molecules through surface anchoring [
17], and changing the incident beam to collimated and focused depolarized laser beam [
18,
19]. Recently, Usui et al. reported that the nonlinear optical reorientation of oligothiophene-doped PSLC systems could be applied to bendable optical limiters [
20]. Therefore, more applications using PSLC systems have been under investigation, such as the fabrication of a microlens array.
There is a growing interest in producing more compact optoelectronic systems that require more investigation to deliver more miniaturized optical elements. A microlens array is a two-dimensional array consisting of small lenses whose diameters are about a millimeter or smaller. Many research scientists have used melting mechanisms to prepare patterned microlens materials through photoresist [
21], wet etching [
22], and laser pulse ablation [
23]. Microlenses play an essential role in many fields, including biomedical imaging systems [
24], optical communication [
25], and 3D imaging [
26]. Therefore, research has been implemented in recent decades to improve the optical properties of a microlens array through surface topographical changes under the application of an external stimulus [
27,
28].
In this work, we establish an innovative approach for fabricating a microlens array with a tunable surface topography and focal length under heat treatment. In our group, a microlens array is fabricated by the light intensity-dependent molecular reorientation of LCs with subsequent photopolymerization [
29,
30]. However, the fabricated microlens array does not possess focal length tunability, and the surface topographical structure change under heat has not been explored. Therefore, the microlens array in this study is fabricated through the photoinduced molecular reorientation of oligothiophene dye-doped LCs embedded in a polymer network. The fabricated microlens array has two unique properties, (i) surface topographical control and (ii) tunable-focus control, all of which occurred due to the nematic-to-isotropic phase transition of the host LC. The results of this study will be helpful in practical applications in providing a microlens array with thermally tunable focus control in the future.
2. Materials and Methods
2.1. Materials
The constituent chemical structures used in this study are shown in
Figure 1. A nematic host LC, 5CB (4-cyano-4′-pentylbiphenyl), was obtained from Merck Ltd., Tokyo, Japan. The oligothiophene dye molecule, TR5 (5,5″-bis-(5-butyl-2-thienylethynyl)-2,2′:5′,2″-terthiophene), was synthesized as previously reported [
7]. The photopolymerizable monomer, A6CB (4ʹ-[6-(acryloyloxy)hexyloxy]-4-cyanobiphenyl), was synthesized according to a previously reported procedure [
31]. The crosslinker, HDDMA (1,6-bis(methacryloyloxy)hexane), was purchased from Tokyo Chemical Industry Co., Ltd., Tokyo, Japan. The photoinitiator, Irgacure 651 (2,2-dimethoxy-2-phenylacetophenone), was obtained from Tokyo Chemical Industry Co., Ltd., Tokyo, Japan, and used without further purification for photopolymerization. A silane coupling agent, octadecyltrimethoxysilane, was purchased from Tokyo Chemical Industry Co., Ltd., Tokyo, Japan. Tetrahydrofuran (THF with stabilizer, 99.5%) and 2-propanol (IPA, 99.5%) were purchased from FUJIFILM Wako Pure Chemical Corp., Osaka, Japan.
2.2. Sample Preparation
First, 0.50 g of 5CB containing 0.10 mol% of TR5, 0.10 g of A6CB, 17.5 mg of HDDMA, and 2.0 mg of Irgacure 651 were added to a small brown glass vial and dissolved in a THF solution. The mixture was magnetically stirred for 1 h until fully dissolved. The solvent was later evaporated in a vacuum desiccator for 2 h. The chemical composition ratio of 5CB, A6CB, HDMMA, Irgacure 651, and TR5 was 84.3, 12.2, 2.87, 0.33, and 0.3 mol%, respectively. The mixture was subjected to a differential scanning calorimeter (DSC, Hitachi High-Tech Corp., DSC7000X, Tokyo, Japan) at a scanning rate of 1 °C/min, and the third scan was reported and shown in
Figure 2. The mixture showed great miscibility with 5CB. The nematic-to-isotropic phase transition temperatures (
TNI) of the mixture were determined as 37.2 °C and 37.1 °C with enthalpy changes (Δ
H) of 0.17 and 0.18 kJ/mol for the heating and cooling processes, respectively.
Second, LC glass cells were prepared using the following steps. Two commercially available glass substrates (2.5 cm × 2.5 cm) were ultrasonically washed in IPA for 30 min. The glass substrates were then treated with a UV-ozone cleaner (NL-UV42, Nippon Laser & Electronics Lab. Co. Ltd., Nagoya, Japan) for 10 min. The cleaned glass substrates were treated with a silane coupling agent by immersing the glass substrates in a 200 mL ethanol solution containing 0.4 g of silane coupling agent for 30 min. The surface-modified glass substrates were then annealed to 120 °C for 2 h to yield glass substrates modified with a silane coupling agent. To fabricate an LC glass cell, two surface-modified glass substrates were bonded with 100 μm thick polyimide tapes.
2.3. Microlens Array Fabrication
The fabrication of a microlens array is shown in
Figure 3. First, the mixture was injected into the prepared LC glass cell at 50 °C by capillary forces. The sample was allowed to rest overnight to stabilize the homeotropic (out-of-plane) molecular ordering of the molecules induced by the silane coupling treatment. The observation of the sample with a conoscopic polarized optical microscope (POM, BX50, Olympus Corp., Tokyo, Japan) equipped with a 546 nm interference filter confirmed the initial homeotropic molecular orientation, indicated by a distinctive isogyre. The homeotropically aligned sample was irradiated with a linearly polarized direct diode (DD) laser beam with a wavelength of 488 nm (EXLSR-488C-200-CDRH, Spectra-Physics, MKS Instruments, Inc., Milpitas, California, USA). The optical setup used in this study is shown in
Figure 4a. The sample was irradiated at 10 W/cm
2 for 10 min and was consistently repeated several times to make a microlens array. After confirming the microlens array by orthoscopic POM (BX50, Olympus, Tokyo, Japan) equipped with a 546 nm interference filter, photopolymerization was performed by irradiating the LC cell with an ultraviolet (UV) light at a wavelength of 365 nm from a UV-LED light source (LHPUV365/2501, Iwasaki Electric Co., Ltd., Tokyo, Japan) at 1.0 mW/cm
2 for 1 h.
2.4. Surface Topographical Characterization
The fabricated LC cell containing the microlens array was carefully opened and analyzed with a confocal laser microscope (LEXT OLS5000, Olympus Co., Ltd., Tokyo, Japan) equipped with a temperature controller stage (HCS402-MK1000TI, INSTEC, Inc., Boulder, Colorado, USA) to characterize the surface topographical change at various temperatures. The height of the microlens array was measured by the confocal laser microscopy and the width of the microlens array was determined through image analysis (ImageJ 1.50i, Wayne Rasband, National Institutes of Health, Bethesda, MD, USA). The height and the width were estimated by taking the average of 12 different microlenses at each temperature with their respective standard deviations.
2.5. Focal Length Measurement
The focal length of the fabricated microlens array was measured experimentally under POM equipped with a hot stage (Tokai Hit, TP-S-100, Tokyo, Japan) by removing the analyzer, sending a linearly polarized light at 90°, and increasing the temperature. First, the base of the microlens array was focused and this distance was denoted as L = 0 μm. The stage was adjusted until the irradiated light was focused on the microlens array, and the distance moved was defined as the focal length. Then, the temperature was increased by 1 °C each time, and the stage was adjusted until the light was again focused. At each temperature, a new focal length was determined and recorded.
3. Results and Discussion
The mixture showed a uniform and transparent yellow color in the LC cell owing to the dopant of the guest dye, as shown in
Figure 5a. A microlens array was fabricated through irradiation with a laser beam at a light intensity of 10 W/cm
2 for 10 min at each spot. The fabricated microlens array could be observed with the naked eye, as shown in the inset of
Figure 5a, and was distributed in a pattern structure with equal distance in the x and y directions. The LC cell containing the microlens array was observed under POM, as shown in
Figure 5b. Following photopolymerization, the surrounding areas of the microlens array showed a polydomain structure due to molecular disordering of 5CB, indicated by different birefringence, as shown in the green, yellow, and red areas. The surface topography of the microlens array was evaluated by carefully opening the LC cell and placing the glass substrate containing the microlens array in a confocal laser microscope.
The 3D and top views of the surface topographical structure of the microlens array are shown in
Figure 6 and
Figure 7, respectively. At room temperature, the microlens array had a mountain-shaped topographical structure with a non-uniform size. As the temperature was increased above the
TNI of the host 5CB, the area surrounding the microlens array became lighter. This suggests that the 5CB molecules encapsulated within a polymer network were transitioning from the nematic phase to the isotropic phase. Interestingly, the microlens also became smaller and more uniform in shape with the increasing temperature. Additionally, as soon as the sample was cooled down to room temperature, the surface topographical structure of the microlens array returned to its original state, showing shape reversibility.
The microlens height and width at different temperatures are summarized in
Figure 8. The microlens height was estimated using the height profile plots provided by the confocal laser microscopy, and the microlens width was measured using ImageJ. The microlens array had an average height of 2.85 ± 1.4 μm and 4.41 ± 2.0 μm at 24 °C and 40 °C, respectively. The microlens array had an average width of 54.6 ± 18.4 μm and 29.6 ± 4.5 μm at 24 °C and 40 °C, respectively. The microlens height increased by roughly 35%, while the width decreased approximately by 45%. Although the microlens height remained almost unchanged, the microlens width became more uniform at higher temperatures. The increase in height and decrease in width might have been due to the thermal expansion of the polymer network and the nematic-to-isotropic phase transition of 5CB in the microlens array, respectively.
The microlens array was formed through the photoinduced molecular reorientation of 5CB when irradiated with a Gaussian beam. The photoinduced molecular reorientation, which was proportionally related to the change in refractive index, was observed through the appearance of concentric diffraction rings, as shown in
Figure 9a. Before the irradiation with a laser beam, the 5CB molecules were homeotropically (out-of-plane) aligned. After increasing the light intensity of the laser beam above a certain threshold intensity, the 5CB molecules began to align parallel to the polarization direction of the laser beam, inducing a homogeneous (in-plane) alignment of the molecules. The direction of molecular alignment of 5CB usually returns to that of the original homeotropic alignment when the light source is removed from the sample. However, the homogeneous alignment became permanent when the sample was irradiated with a laser beam for 10 min, stabilizing the molecular reorientation of 5CB.
The stabilized molecular reorientation of 5CB was probed with a 633 nm linearly polarized He-Ne laser beam (Melles Griot, 05 LHP 151, Pneum, Saitama, Japan), where the sample did not show any absorbance at that wavelength, showing memorized concentric diffraction rings, as shown in
Figure 9b-1. The memorized diffraction rings did not possess any polarization dependency stating that the microlens was fabricated through thermal distortion, i.e., order-to-disorder molecular reorientation. In other words, the center of the microlens showed the molecular disordering of 5CB, while the outer section showed the molecular ordering of 5CB. This was due to the Gaussian light intensity distribution of the laser beam, where the center of the laser beam had greater light intensity than the outer side. Therefore, the 5CB molecules within the center of the irradiated area remained isotropic (disordering), while the 5CB molecules located in the outer region of the laser beam remained homogeneously aligned.
The molecular reorientation of 5CB was stabilized through the photopolymerization of monomers inside the irradiated area as the sample was exposed to high light intensities for long periods, triggering the photoinitiator to initiate the photopolymerization. Owing to the monomer concentration difference between the irradiated and unirradiated areas with the laser beam, the monomer diffuses from the unirradiated area to the irradiated area, causing a high content of polymers inside the irradiated region [
28,
29]. This caused the microlens array to gain a mountain-shaped topography due to the polymer accumulation inside the microlens. Hence, the microlens array consisted of disordered 5CB molecules with an accumulation of polymers. The fabricated microlens array was further stabilized by irradiating with a UV light source at 1 mW/cm
2 for 1 h across the whole area. The molecular reorientation of the microlens was again probed with the He-Ne laser beam, as shown in
Figure 9b-2.
The focal length is an essential property of a microlens array, and different approaches to tune the focal length through an external stimulus were explored. The focal length of the microlens array in different temperatures was measured (
Figure 10). At room temperature, the focal length of the microlens array was 640 μm and remained constant until 30 °C because the 5CB molecules remained in the nematic phase. As the temperature was increased to 31 °C, the focal length slightly decreased until it reached a plateau of 634 μm at 34 °C. This decrease in focal length was due to the surface topographical change in the microlens array and the nematic-to-isotropic phase transition of 5CB. The focal length gradually increased when the temperature reached 35 °C and remained unchanged after 36 °C. After 37 °C, the molecules in the microlens array became isotropic and the surface topography remained roughly unchanged. However, the focal length became smaller compared to the focal length at room temperature because 5CB in the microlens array became isotropic. The microlens array was cooled to room temperature and the focal length was reversible. This is comparable to the reversibility of topographical change under heat treatment, which also induces the change in focal length. Even though the microlens only showed the focal length tuning by 8 μm, this could be useful to fabricate thermally tunable microlenses in the future.