1. Introduction

In recent years, micro/nano lens arrays (M/NLAs) have attracted significant attention because of their broad range of applications, including enhancing light-coupling efficiency in organic light-emitting devices (OLEDs) [1], photovoltaic cells [2,3], photodetection [4,5], three-dimensional (3D) imaging [6,7], microfluidic devices [8], biomedical imaging [9,10], and artificial compound eyes [11]. Particularly, NLAs show a unique ability of being able to overcome optical diffraction limits to achieve super-resolution near-field imaging and enhanced optical signals even in low light. Therefore, NLAs have shown potential in applications such as super-resolution imaging [12], super metalenses [13,14], surface-enhanced Raman scattering (SERS) [15], and nanophotonics [16,17]. Micro/nanolenses are core optical elements that can considerably reduce the size of the device and improve its integration. In addition, the performance of the device can be significantly improved by the micro/nanoscale characteristics. Similar to the parameters of conventional lenses, the two most important parameters of micro/nanolenses are the focal length (f) and numerical aperture (NA). For some typical applications such as optical interconnects, two-dimensional (2D) high-density optical sensor arrays, laser protection goggles, and 2D optical data-recording and projection systems [18,19], large NAs are imperative to achieve high signal-to-noise ratios. The manufacturing methods for micro/nanolenses should facilitate good material adaptability and the control of the aspect ratio in order to control the lens shape.

Several approaches have been employed for manufacturing high-quality M/NLAs, such as laser-induced writing [20–23], two-photon polymerization [24], and lithographic transfer [25]. These are straightforward routes of engraving a customized 3D pattern onto a material, and can yield MLAs with superior aspheric surface shapes. Laser-induced writing enables good material adaptability; nevertheless, high laser quality and precise energy control are required to avoid manufacturing defects. Two-photon polymerization is only suitable for photosensitive materials, and the selection of materials is limited. In addition, both techniques are expensive and incur trade-offs between accuracy and efficiency. Lithographic transfer is widely used and is suitable for the mass production of microlens arrays. However, it cannot be adopted for different processing requirements where the mask needs to be redesigned. In addition, other techniques, such as polymer swelling [26,27] and thermal reflow [28] have been used to fabricate MLAs with tunable surface profiles. These techniques are flexible in application and low in cost; however, it is difficult to control the geometric shape of lenses and the uniformity of the lens arrays (LAs) using them. For sub-microlens or nanolens arrays, methods such as laser ablation [29], DNA-origami-based self-assembly [30], self-organized polymer dewetting [31], and electron-beam lithography (EBL) [14] have been explored. In addition to EBL, other technologies have the disadvantage of poor controllability. EBL enables extremely high manufacturing accuracy; however, considerably expensive equipment is required and processing efficiency is quite low. Consequently, there is an urgent demand for an approach to fabricate MLAs and NLAs with tunable lens curvatures and sizes that range from micrometers to nanometers, consecutively on different substrates.

Inkjet printing is a direct-additive and maskless fabrication method [32–34], which has demonstrated good flexibility in the fabrication of MLAs on various substrates in numerous micro-optical systems. This method is widely used, and it has no special requirements for substrate materials. To avoid nozzle blockage, the viscosity of the ink material should be between 1 and 50 cps; this also limits the selection of different refractive index materials. In 2007, Park et al. introduced electrohydrodynamic jet (E-jet) printing as a viable technique for the fabrication of structures at the micro/nanoscale, with various applications related to the printing of different materials [35]. A new print head design was demonstrated by Hashimdeen et al. for high-resolution E-jet printing [36]. Additionally, E-jet printing has been used to fabricate micro/nanodevices with high resolution [37–40]. However, the cross-scale fabrication of MLAs and NLAs with optimal parameters remains a challenge, in terms of the implementation method for stable printing to achieve the best morphological parameters of LAs, and the flexibility of the approach to adjust the variable optical parameters of LAs.

In this paper, we report the one-step maskless fabrication of lens arrays from the microscale to the nanoscale via E-jet printing. The stable cone-jet printing mode with optimized parameters was implemented for realizing the best morphological parameters of the M/NLAs. The optical parameters (f and NA) of the M/NLAs were flexibly adjusted via the simple wettability modification of the substrate. By applying UV-curable adhesives, we realized the fabrication of M/NLAs on glass and silicon substrates. Further, NLAs with various parameter values were then straightforwardly fabricated on an atomic force microscopy (AFM)-probe cantilever and graphene. The lens-modified AFM probe can be used in applications where synchronous super-resolution imaging is required. The NLA-modified graphene can be used in photoelectric-sensing applications. Clear images were obtained in the optical analysis, demonstrating the excellent optical performance and good homogeneities of the LAs.

2. Materials and methods

The wettability of the substrate and the viscosity of the UV-curable adhesive dominate the morphologies of M/NLAs. Therefore, the substrates must be treated to adjust the contact angles (CAs). We adopted four kinds of UV-curable adhesive to observe the influence of viscosity of the adhesives on the optical parameters of M/NLAs. As optical components, the morphologies and homogeneity of M/NLAs are very crucial for the application. In our experiments, we used several tools to measure and characterize the M/NLAs.

2.1. Surface modification

Glass (150-µm thick) and silicon (500-µm thick) were selected as the substrates (20 mm × 20 mm plane size, purchased from Sigma, Inc.). Firstly, the glass and silicon substrates were successively ultrasonically cleaned by immersion in acetone (purity, 99.9%) and ethanol (purity, 75%) at 40 °C and 40 kHz for 30 min. Subsequently, the substrates were rinsed with deionized (DI) water, and dried under nitrogen. The wettability of the substrate surface was modified by enhancing their hydrophilicities by treatment with O₂ plasma, while their hydrophobicities were enhanced by treatment with PDMS, as reported in our previous research [41]. The CAs of the substrates could be adjusted from 14 ± 2.2° to 115 ± 4.3°. Monolayer graphene was prepared by the Institute of Metal Research (IMR), Chinese Academy of Sciences using a CVD growth method, as previously reported [42].

2.2. Materials of UV-curable adhesives

The sharp microcapillary nozzles were chosen as the printing nozzles with the inner diameter ranging from 0.2 to 30 µm, which were manufactured using glass capillaries pulled via a pipette puller (P-97, Sutter, Inc., USA). The UV-curable adhesives (purchased from Thorlabs, Inc.) were used as printing inks. The detailed parameters are shown in Table  1. After printed on the substrate, the adhesive micro-nano droplet arrays were exposed to 355 nm ultraviolet (UV) light at 1000 mJ/cm². After 10s of UV exposure for precuring the micro-nano droplets, the M/NLAs can completely be solidified after the treatment under 30 min of UV exposure.

Table 1. Parameters of UV-curable adhesives

View Table

2.3. Characterization

Scanning electron microscopy (SEM; Quanta 400 FEG) was used to analyze the morphologies of M/NLAs. The 3D morphologies were measured in the tapping mode by employing AFM (Dimension Icon, Bruker Nano, Inc., USA), using a TESP-V2 probe with the tip curvature radius of 8 nm. The uniformity of large-scale M/NLAs was characterized using an optical microscopy (HIROX KH-7700) together with a charge-coupled device (CCD) video camera (Point Gray, Canada).

3. Results and discussion

3.1. Fabrication of M/NLAs using E-jet printing system

The schematic of the approach used to fabricate the M/NLAs by E-jet printing is displayed in Fig. 1. As shown in Fig. 1(a), a DC voltage is applied between the nozzle and the substrate. The electric field at the nozzle tip results in a concentrated charge on the pendant drop that emanates from the tip. In our experimental system, the dual inclined CCD cameras and a PZT nanopositioner are used to realize spatial imaging of the fabrication state and to adjust the relative position between the tip and the target location on the substrate.

The concrete process for the adjustment of the distance between the tip and the substrate is similar to the reported details in our previous study [43]. In this case, a low trigger voltage can be used to avoid satellite drops induced by electrohydrodynamic instability. The shear stress generated by the concentrated charge forces the meniscus-shaped drop to become conical in shape, which is referred to as a “Taylor cone”. A droplet is released from the nozzle tip when the ink surface tension is overcome by the shear stress generated by the charge, the parameters of droplets are controlled by the basic E-jet-printing parameters (including the printing mode, inner diameter of the nozzle, and amplitude and duration of applied voltage) and the movement of the translation stage to produce MLAs and NLAs with controllable sizes. Fig. 1(b) shows the schematic of the two basic E-jet-printing adjustment processes, namely, location distribution and printing modes. The location distribution of the M/NLAs (i.e., the gaps between the lenses), which can be controlled by the programed movement paths of the translation stage. The inset shows the key printing modes: micro dripping and stable cone-jet, which lead to lens surfaces of different smoothness. Droplets with accurate volumes can be precisely deposited at specified positions according to the precise path programmed and the designed digital pattern. Meanwhile, as shown in Fig. 1(c), when combined with the adjustable E-jet-printing-system parameters, including voltage, and inner-nozzle diameter, the printing times for the UV-curable adhesives can be precisely controlled, which leads to a variety of profiles and characteristics for individual lenses; hence LAs with various size and fill factors can be realized. Fig. 1(d) and 1(e) show the schematics of the curvatures of lens that can be adjusted by controlling the wettability of the substrate and the viscosity of the UV-curable adhesive. As the SEM image in Fig. 1(f) shows, homogeneous MLAs with an average diameter of 20 µm are prepared. MLAs with four sizes that range between 33.5 µm and 2.5 µm (average diameters) are shown in Fig. 1(g), and the inset shows the magnified images of the MLAs with average diameters of 2.5 µm and 5.25 µm. The AFM scanned images presented in Fig. 1(h) show the 3D topographies of the “NANO LENS” and “SIA” letter patterns, which are composed of 1016 nanolenses with an average diameter of 824 nm, and 28 microlenses with an average diameter of 3.56 µm.

 

Fig. 1. Schematic of the experimental process used to controllably and adjustably fabricate M/NLAs. (a) Schematic of the E-jet printing system used to fabricate the MLAs and NLAs. (b) Schematic of two basic E-jet-printing adjustment processes: location distribution and printing modes. (c) Schematic showing different LAs with adjustable sizes and fill factors. (d) Schematic showing lens curvature adjustment by controlling substrate wettability and the viscosity of the UV-curable adhesive. (e) Schematic showing LAs with adjustable curvatures. (f) SEM image of an MLA array showing homogeneously sized lenses, with a magnified image shown in the inset. (g) SEM image of an array with diverse MLA sizes, with a magnified image of two small lenses shown in the inset. (h) AFM-scanned image showing the 3D topographies of “NANO LENS” and “SIA” letter patterns composed of NLAs and MLAs, respectively.

Download Full Size | PDF

3.2. Analysis of the printing modes and adjustment of the lens curvature

When a voltage is applied between the tip and the substrate, the electric force suffered by the liquid pendant is sufficient to overcome the surface tension and eject the liquid. Based on the mechanism associated with electrohydrodynamic instability, the liquid can be ejected in multiple ejection modes, including dripping, micro-dripping, stable cone-jet, unstable cone-jet, and multi-jet [44,45]. As the schematic of the major ejection modes shown in Fig. 2(a), the state of the ejected liquid ranged from the dripping or micro-dripping mode to the multi-jet mode with increasing electric-field intensity during the experiments. As shown in Fig. 2(b), the distance between the tip and glass substrate was maintained at 10 µm, and multiple ejection modes can be produced by varying the inner diameter of the tip and the voltage. The jet-mode domains for print nozzles with different inner tip diameters were controlled by varying the applied voltage during the E-jet printing of the NOA 63 UV-curable adhesives. As a consequence, the voltage needed to be increased as the nozzle diameter was increased in order to maintain a specific ejection mode. In contrast, the ejection mode generally varies from the dripping mode to other modes with increasing voltage at a fixed inner nozzle diameter. For instance, the liquid-ejection mode varied from micro-dripping to stable cone-jet as the voltage was increased from 0.25 to 0.45 kV for an inner nozzle-tip with diameter of 10 µm.

 

Fig. 2. Analyzing various jet printing modes and adjusting the lens curvature. (a-b) Schematic illustrations and analyses of the major printing modes used in E-jet printing. (c) 2D AFM images of micro lenses fabricated by the E-jet printing modes of dripping, micro-dripping, stable cone-jet, unstable cone-jet and multi-jet modes, respectively. (d) RMS roughnesses of the top surfaces of lens, and NSDs of lens diameter as functions of inner tip diameter in five jet printing modes. (e–h) 3D AFM images of micro lenses fabricated on substrates with CAs of 28°, 50°, 75°, and 108°, respectively. (i) AFM scanned profiles of the paths marked in panels (e–h), respectively. (j–m) 3D AFM images of micro lenses printed with UV-curable adhesives with viscosities of 300, 1200, 2000, and 5000 cps, respectively. (n) AFM cross-sections of the paths marked in panels (j–m).

Download Full Size | PDF

Smoothness and uniformity, which are the two key indexes for determining the surface quality of an optical lens, are important for ensuring the high performance of an MLA. Consequently, we examined various E-jet printing modes in detail to gain insight into the influence of various printing parameters on lens quality. The AFM images of microlenses fabricated in the dripping, micro-dripping, stable cone-jet, unstable cone-jet and multi-jet modes are displayed in Fig. 2(c). In order to investigate the influence of ejection mode on MLA quality, we statistically analyzed the root-mean-square (RMS) roughness of each lens surface, and the normal standard deviation (NSD) of the lens diameter, as shown in Fig. 2(d), the NSD is the ratio of the standard deviation to the average diameter of the lenses. For the tip with an inner diameter of 10 µm, the pulsed droplet varied from the dripping to the multi-jet mode with increasing voltage. The RMS roughnesses of the nanoparticle dots corresponding to the five modes were found to be 21.3, 12.7, 9.1, 3.8, and 2.5 nm; meanwhile, the lowest diameter NSD was observed in the stable cone-jet mode. Although lenses with the smoothest surface were fabricated in the multi-jet mode, it was unsuitable for the fabrication of lenses with stable sizes and locations. The nonuniformity of lenses was exacerbated by the electrohydrodynamic instability with increasing voltage. To improve the controllability and consistency of fabrication of LAs in large scale, the stable cone-jet mode is optimal for fabricating LAs with the best uniformity. Compared to lithography, the surface smoothnesses of the MLAs prepared by E-jet printing are better able to meet the requirements of numerous applications in which rigorous optical MLA performance is indispensable.

In addition, the curvatures of the lenses are mainly dominated by the wettability of the substrate. Meanwhile, the ink viscosity also has a significant impact on the curvatures of the lenses. Based on the results described above, an AC voltage of 0.5 kV was applied under the stable cone-jet printing mode. The duration for printing each microlens was 1s, the inner diameter of the nozzle tip was 10 µm. Lenses fabricated using UV-curable adhesives with different viscosities on substrates with different wettabilities were next investigated. The AFM images and corresponding scanned profiles shown in Fig. 2(e)–2(h) reveal that the CAs of the MLAs fabricated with the NOA 63 UV-curable adhesives on substrates with different wettabilities are 28°, 50°, 75°, and 108°, respectively. As shown in Fig. 2(i), the volume, diameter, height, and diameter-to-height ratio (d/h) of the lenses vary from 99.3 to 102.9 fL (with 3.6% relatively deviation), 17.86 to 9.19 µm, 0.79 to 2.77 µm, and 22.62 to 3.32, respectively, as the CA is increased from 28 to 108°, respectively. Fig. 2(j)–2(m) display the AFM images of MLAs printed with UV-curable adhesives of the NOA 61, NOA 65, NOA 63, and NOA 68 UV-curable adhesives on a substrate with a CA of 93°, respectively. Fig. 2(n_ reveals that the volume, diameter, height, and d/h ratio of the lenses vary from 101.2 to 12.6 fL, 10.21 to 6.36 µm, 2.31 to 0.78 µm, and 4.42 to 8.15, respectively, as the viscosity of the UV-curable adhesive increases from 300 to 5000 cps.

3.3. Fabrication of MLAs and statistical analysis of the various parameters

According to the above results, a variety of microlens patterns can be produced by precisely tuning the viscosity of the UV-curable adhesives and the wettability of the substrates. Fig. 3(a)–3(e) shows the arrays with microlenses of different diameter. The average diameters of the microlenses shown in Fig. 3(a–c) are 5.66 ± 0.27 µm, 7.52 ± 0.34 µm, and 3.46 ± 0.16 µm, the deviations of the lenses diameters are less than 5%. Moreover, E-jet printing can also fabricate more complex structures, such as arrays of microlenses with gradually increasing size and desired letter patterns, such as “SIA”, as shown in Fig. 3(d) and 3(e). The corresponding 3D morphologies and AFM cross-sectional profiles are shown in Fig. 3(f)–3(j), which confirm uniformity and smoothness. Large-scale MLA images were obtained by SEM, as shown in Fig. 3(k)–(o). For average diameters of 20.2 µm, 11.8 µm, 17.4 µm and 5.25 µm, the corresponding average heights of 3.08 µm, 1.79 µm, 2.62 µm, and 0.81 µm, respectively, and the average center gaps are 40 µm, 20 µm, 40 µm, and 40 µm, respectively. E-jet printing can be used to create size-variable MLA patterns by dynamically modifying the voltage. Fig. 3(o) shows a large MLA composed of four homogeneous sub-MLAs of different size; each homogeneous sub-MLA consists of 100 (10 × 10) microlenses with average diameters of 1.21 µm, 1.23 µm, 1.51 µm, and 3.15 µm, average heights of 0.15 µm, 0.19 µm, 0.27 µm, and 0.68 µm, and average center gaps of 10 µm each.

 

Fig. 3. Precisely controlling size during the fabrication of an MLA by E-jet printing with various experimental parameter values. (a–e) AFM images of MLAs with different sizes, gaps, and curvatures. (f–j) 3D morphologies and cross-sectional profiles corresponding to panels (a–e), respectively. (k–n) Large-scale top-view SEM images of various MLAs, the insets show magnified images tilted at 45°. (o) Top-view SEM image of a large MLAs composed of four homogeneous MLAs with different lens sizes.

Download Full Size | PDF

As mentioned above, several parameters influence the topography and size of an MLA, including the inner tip diameter, voltage, printing time, substrate wettability, and the viscosity of the UV-curable adhesive. Meanwhile, the topography and size of the MLA determines its final optical performance, which is mainly influenced by the NA, the focal length (f), and the surface smoothness.

On the basis of geometry and optical theory [46,47], the focal length (f) can be determined using the following formula:

 

Fig. 4. Relationships between various MLAs parameters and the different fabrication conditions. (a) Diameter and height of lens as functions of tip inner diameter with different applied voltage amplitudes (0.4 kV and 0.8 kV). (b) Diameter and height of lens fabricated by tip with different inner diameters (2 µm and 10 µm) as functions of applied voltage amplitude. (c) Ratio of diameter and height (d/h) and volume of lens fabricated with different durations as functions of the substrate CA. (d) d/h ratio and volume of lens as functions of the viscosity of the UV-curable adhesive. (e-h) Heights, diameters, focal length (f) and NA of lenses with different fabrication conditions as functions of the substrate CA.

Download Full Size | PDF

As shown in Fig. 4(a), the diameters and heights of the MLAs were observed to increase from 1.59 to 32.45 µm and 0.11 to 2.17 µm, respectively, as the inner tip diameter was increased from 1 to 25 µm, respectively. Similarly, Fig. 4(b) reveals that the diameter and height of MLAs increase with increasing voltage.

The relationships between the CAs of the substrates, the viscosities of the UV-curable adhesives, and the d/h ratio at different printing times and different volumes of MLAs were explored, as shown in Fig. 4(c) and (d). We conclude that the d/h ratio of the MLAs decreases with increasing CA. In addition, the CA exerts a stronger influence on the d/h ratio of the MLAs printed with low-viscosity UV-curable adhesives compared to high-viscosity adhesives. Other than the minor experimental deviation, the different CAs of the substrates have negligible effect on the volume of lenses during the fabrication process under the same printing conditions, while the height of lens increases with increasing the CA and the diameter of lens decreases with CA. The volume of MLAs increases with increasing printing time and decreases with increasing viscosity when the values of the other parameters remain unchanged.

MLAs of the same volume were fabricated on glass substrates with different CAs to investigate the relationships between the optical parameters and MLA size and to analyze the influence of the CA on the optical parameters. Lenses 1 and 2, 3 and 4, and 5 and 6, were fabricated by printing the NOA 61 (viscosity 300 cps), NOA 65 (viscosity 1200 cps), and NOA 63 (viscosity 2000 cps) adhesives, respectively. The volumes of lenses 1, 2, 3, 4, 5, and 6 were 94.15 fL, 993.68 fL, 50.86 fL, 482.56 fL, 26.37 fL, and 231.83 fL, respectively. The refractive indices n of NOA 61, NOA 63, and NOA 65 are 1.56, 1.56, and 1.524, respectively. The low values of f and NA calculated by Eqs. (1) and (2) are as expected because small microlens diameters impose small values on the radius of curvature, which results in the collection of small amounts of light. The relationships between the focal length, NA of MLAs with different sizes (see Fig. 4(e) and 4(f) for the statistics of the heights and diameters of the lenses), and the various CAs of the substrates are presented in Fig. 4(g) and 4(h), respectively. The focal length increases with increasing lens diameter d and decreases with increasing CA, while the NA increases with CA. For the present experiment, the f values of the MLAs ranged from 9.15 to 445.68 µm, and their NAs ranged from 0.04 to 0.49 for lenses on glass substrates.

3.4. Characterizations of the optical performance of the fabricated MLAs

To investigate the imaging performance of MLAs, we designed a simple optical-imaging system, as shown in Fig. 5(a). The MLAs was positioned on a stage capable of moving along the optical axis. The transmission mode was adopted and the letters “SIA” were illuminated by white light; As a result, an array of projected “SIA” letters was clearly observed by the CCD camera. First, the focusing capability of the fabricated MLA was investigated and characterized via optical microscopy. An MLA with an average diameter of 15.16 µm, height of 3.56 µm, f of 18.84 µm and NA of 0.40 was used for the focusing test. The brightness image shown in Fig. 5(b), and Fig. 5(c) shows the normalized light intensity distribution of the lenses along the green line in Fig. 5(b). These results of which indicate that the sizes and focal lengths of the MLAs are uniform. As the imaging test of the MLAs shown in Fig. 5(d)–5(g), the average f and NA of MLAs of Fig. 5(d) and 5(g) are 24.37 µm and 0.40, 31.48 µm and 0.40, respectively. The arrays of the projected “SIA” can be clearly observed via the MLAs with different parameters through the CCD camera. According to the magnified image shown in Fig. 5(e) and the inset image of Fig. 5(g), the sharp and high-contrast images demonstrated the good imaging performance and optical property of the MLAs. Meanwhile, Fig. 5(h) and 5(i) show the focusing images obtained from an MLA composed of microlenses of different sizes. The diameter and height of the big lens are 64.02 µm and 10.86 µm, respectively, and the average diameter and height of the three small lenses are 32.70 µm and 5.58 µm, respectively. The corresponding projected “SIA” images move in and out of focus as the MLA is moved along the optical axis. The clear focusing images of projected “SIA” observed via the big lens and small lenses are shown in Fig. 5(h) and 5(i), respectively. The f and NA values of big lens are 100.33 µm and 0.32, the average f and NA values of small lenses are 51.04 µm and 0.32. It reveals that the sizes of the microlens clearly influences the focal length of the MLA. Therefore, it is necessary to precisely control the values of the E-jet-printing parameters to fabricate the target microlenses. Furthermore, it is also possible to produce MLAs with various focal lengths for 3D optical imaging. Based on these results, we can conclude that the E-jet printing can be utilized to fabricate high-quality fabrication of MLAs with uniform or variable parameters on the same substrate using a reliable and flexible fabrication process.

 

Fig. 5. Experimental MLA-imaging performance testing. (a) Schematic of the MLA-imaging test system. (b-c) Focusing capability test of the fabricated MLAs. (d-e) Image of the letters “SIA”. focused by MLAs. (e) show the magnified image of the area enclosed by the red dashed line in (d). (f) SEM images of MLAs arranged in the “SIA” letter pattern. (g) Image of the letters “SIA” as focused through each lens of the “SIA”-patterned MLAs shown in (f), inset image shows the high magnification image projected through a single lens, the inset scale bar is 5 µm. (h-i) The images of the letters “SIA” as focused by micro lenses with different sizes.

Download Full Size | PDF

3.5. Cross-scale fabrication of lens arrays from microscale to nanoscale

Compared to microlenses, nanolenses provide several unique advantages for a wide range of nanotechnology applications. However, it is quite challenging to fabricate NLAs via E-jet printing. On one hand, electrohydrodynamic instability leads to the generation of satellite droplets, while on the other hand, the droplet fusion caused by the Ostwald ripening effect must be overcome [48]. To reduce the formation of satellite droplets, we used a lower starting voltage in our experiments, while the distance between the tip and the substrate must be maintained to within approximately 1-5µm during the printing process to prevent droplet fusion, which increases the risk of breakage of the nozzle tip due to contact with the substrate. To resolve this issue, dual-CCD cameras in the tilting-observation mode were adopted to determine the distance between the tip and the target printing position of substrate, and a piezoelectric ceramic driver was used to precisely control the position of the substrate. The viscosity of the adhesives is also an important factor when fabricating the NLAs. Therefore, we printed NLAs on various substrates with UV-curable adhesives with different viscosities. First, the NOA 63 UV-curable adhesive was adopted, the voltage was set as 0.15 kV, the inner diameter of the nozzle tip was 200 nm, the printing time was 1 s, and the CA of the substrate was 75°. As shown in Fig. 6(a), with the imitation of graphene lattice structure, a pattern of hexagon array consisting of seven nanolenses was printed on the silicon substrate. The average diameter and height of the nanolens were 119.7 nm and 14.3 nm, respectively, and the volume of a single droplet was calculated to be 8.19 × 10⁻⁵ fL. To the best of our knowledge, this is the smallest droplet volume achieved through E-jet printing to date.

 

Fig. 6. Controllable fabrication of large scale NLAs by E-jet printing. (a–d) AFM images of NLA patterns with different sizes and gaps fabricated on silicon substrate. (e) AFM image of the NLA patterned with the letters “NANO LENS”. (f–j) 3D AFM topography maps corresponding to the NLAs in panels (a–d), respectively, and scanned profiles. (k–o) SEM images of NLAs and MLAs with different sizes and gaps fabricated on the cantilever of an AFM probe. The insets in panels (l–n) show magnified images of the areas enclosed by the blue, red, and green dashed lines in panel (k). (p–r) AFM images of NLA patterns with different sizes and gaps fabricated on graphene. (s) SEM image of the NLA fabricated on graphene depicted in panel (r). (t) AFM scanned profiles corresponding to panels (p–r).

Download Full Size | PDF

Next, the NOA 61 UV-curable adhesive was used for fabricating the nanolens. As shown in Fig. 6(b)–6(d), the voltage was set as 0.18 kV, the inner diameter of the nozzle tip was 1 µm, and the average printing time of single nanolens in each array were 0.02 s, 0.05s, and 0.08s, respectively. We fabricated three nanolens arrays with whole time-consuming of 60s, which contains nanolenses of 900 (30 × 30), 400 (20 × 20), 225 (15 × 15) with different gap spacings on silicon substrate, respectively. The average diameters of the NLAs were 384 nm, 580 nm, and 714 nm, and their average heights were 73 nm, 109 nm, and 132 nm. The center-spacing gaps of these lenses were 0.8 µm, 1 µm, and 1.6 µm, and their margin-spacing gaps were 420 nm, 420 nm, and 890 nm. This reveals that it is difficult to fabricate NLAs with small margin spaces when printing a low-viscosity adhesive. This can be ascribed to droplet fusion at the nanoscale, which is more exacerbated for nanodroplets with low viscosities. To realize NLAs with margin spaces (sub-200 nm) smaller than visible wavelengths [49], which is significant for superlens applications such as super-resolution imaging, we used the NOA 65 UV-curable adhesive for fabricating LAs with complex shapes such as “NANO LENS” lettering. As shown in Fig. 6(e), we successfully fabricated a lens array containing 1016 nanolenses on an area of 2000 (40 × 50) µm². The center-spacing gaps and margin-spacing gaps of these lenses were 1 µm and 160 nm, respectively. To further demonstrate the controllability of the method, we used the NOA 68 UV-curable adhesive to print an NLA on an AFM probe cantilever (TESP-V2, Bruker, Inc.) that was only 42 µm wide, and where considerably finer control was required during printing. As shown in Fig. 6(l)–6(n), the average diameters of the NLAs were 936 nm, 698 nm, and 445 nm, respectively, with a consistently lens center spacings of 2 µm. Fig. 6(o) shows MLAs with spacing of 10 µm, 8 µm, and 5 µm printed onto the same cantilever beam.

To further broaden the applications of NLAs, we used the NOA 63 UV-curable adhesive to conduct experiments on graphene surfaces. Fig. 6(p)–6(t) show NLAs fabricated on graphene surfaces. The average diameters of the NLAs were 942 nm, 836 nm and 596 nm, the average heights of the NLAs were 118 nm, 95 nm, and 68 nm, and the center-spacing gaps of these lenses were 3 µm, 5 µm, and 3 µm, respectively. The insets in Fig. 6(q)–6(t) show high resolution images and the corresponding scanned profiles of the graphene, which has an average height of 1.2 nm.

4. Conclusion

In this paper, we present a simple, maskless, and effective method for the controllable and straightforward fabrication of M/NLAs on various substrates using E-jet printing. Homogeneous, smooth, NA-variable LAs were flexibly fabricated with nanoscale-to-microscale sizes on a variety of substrates with the large scale, the diameters of which range from 64 µm to 120 nm, and NAs in the range of 0.04-0.49. The fabrication method developed herein is potentially applicable to multiple micro/nano optical-research fields, such as flexible curved MLAs, artificial compound eyes, and large-scale nano metalenses for super resolution imaging. In particular, the NLAs fabricated on an AFM probe are not only usable as movable superlenses for super resolution imaging and nanojet lithography on the large scale, but they can also act as multi-dip pens for nanolithography applications [50]. Meanwhile, the NLA-modified 2D nanomaterials (e.g., graphene and MoS₂), can be used in nanophotonics electronic-sensing devices. The NLAs can be used as nanostructure template for patterning functional nano materials, which could be used for surface enhanced Raman scattering, nanodevices and flexible electronics. In the future, improving the aspect ratio and realizing smaller functionalized NLAs are worthy of further research. The techniques developed for fabricating MLAs and NLAs can be applied to manufacture wide and diverse integrated systems of micro-nano optical devices.