1. Introduction

Owing to their good optical transparency and biocompatibility, biological hydrogel materials have emerged as optimal optical materials for structuring biological devices in biophotonics applications [1–5]. Hydrogels endowed with functional properties (e.g., stimulated response) are “tunable and smart” materials [6–9], exhibiting various photonic and optical properties; these materials have attracted considerable attention of researchers in the field of intelligent manufacturing [3,10–13]. Customizable hydrogel-based devices with environment-dependent structure sizes and refractive indices offer many prospects for novel bioengineering applications. Protein-based hydrogels have been increasingly utilized in micro-optical devices [14,15], optical waveguides [5], biosensing [16,17], and bioactuators [18], even in the field of the fiber sensing [19,20]. Remarkably, these novel biological devices are typically fabricated with planar micro-/nanoscale structures. Nevertheless, three-dimensional complex hydrogel structures are more tunable and provide more features for microoptical devices. It would be interesting and significant to make better use of diverse hydrogel structures along with the stimulated response characteristics, for enabling smart biophotonics.

For optical components, it is crucial to reduce chromatic aberration; this need has attracted significant attention and many efforts have been made to improve optical systems correspondingly [21]. In particular, biological photonic devices are typically used in complex environments, which not only include solutions and ions but also often contain biological and polymeric materials and enzymes [22,23]. Thus, apochromatic biological hydrogel-based micro-optical elements are vital and more significant for biological optical applications than traditional optical elements that work in free space [24]. More often than not, monochromatic aberrations (such as defocusing and spherical aberrations) can be easily corrected by designing and enabling aspheric surfaces or freedom optical surfaces [25,26]. However, chromatic aberration is unusual and cannot be corrected because of the refractive index and dispersion differences, which depend on the materials and optical elements [27,28]. Hence, two approaches have been proposed to achieve achromatism: 1) combining materials with different dispersion characteristics or 2) using refractive-diffractive surfaces with different Abbe numbers. Based on this, multicomponent microlens structures combining two different photoresist materials have been reported for obtaining chromatic aberration [29]. Simultaneously combining refractive and diffractive surfaces, two different photoresist materials, have also been demonstrated to drastically reduce chromatic aberration [30]. In general, two types of biological materials are used in optical elements or systems to achieve different dispersions or Abbe numbers. However, it is difficult to create achromatic optical elements in biological materials using the above methods; these materials are typically soft and flexible, and their curability significantly differs from that of photoresists. In addition, adjustable hybrid varifocal diffractive–refractive lenses have been constructed by introducing a tunable liquid crystal layer or dispersion ability to reduce chromatic aberration [31–33]. Nevertheless, to the best of our knowledge, achromatic micro-optical elements have rarely been fabricated on biological hydrogel-based materials. This method, which uses smart tuning, opens up a novel path toward achromatic aberration, which enables achromaticity in biological micro-optical elements by utilizing the stimulated responsive properties of hydrogels.

In this study, a novel biological hydrogel-based achromatic refractive-diffractive micro-optical element (RDMOE) was successfully fabricated using femtosecond laser lithography technology (FsLT) for achromatism reduction. The facile, programmable, non-contact, and maskless fabrication route of the FsLT provided high resolution and acceptable biocompatibility, demonstrating flexible potential for further preparation of diverse biological structures [34–40]. In addition, benefiting from the stimulated responsive property of the hydrogel, pH modulation yielded swelling and refractive index variation in the fabricated structures, enabling the achromatic function. First, we experimentally fabricated a separate hydrogel microlens and Fresnel zone plate (FZP), and measured the tunable focusing performance changes while varying the pH value; the results were consistent with those of simulations. In the second step, we designed a biological hydrogel-based achromatic RDMOE and demonstrated achromatism with respect to three wavelengths using only one material consisting of a microlens and FZP with tunable focusing properties. Subsequently, a biological hydrogel-based achromatic RDMOE was fabricated using the FsLT, and the achromatic aberration with respect to the three wavelengths was characterized. The potential applications of such a hybrid microoptical element include biomedical imaging and optical biology sensing.

2. Experimental details

In our experiments, bovine serum albumin (BSA) was used as the hydrogel material to fabricate optical elements. BSA (concentration, approximately 500 mg/mL) and photosensitizer Rhodamine B (concentration, 2 mg/mL) were combined with a hydrogel photosensitizer ink (HPI), which is based on our previous work [5]. The HPI was drop on the glass cover by a needle and a squared enclosure was constructed with the glue to fill the HPI. The thickness of the HPI is larger than the thickness of fabricated microstructure. HPI-based hydrogel micro-optical elements were directly fabricated on a transparent cover glass using the FsLT. The experimental FsLT system is shown schematically in Fig. 1. A commercial femtosecond fiber laser system was used; it generated a laser beam with a central wavelength of 800 nm, mode-locked duration of approximately 120 fs, and a repetition rate of 80 MHz. The femtosecond laser beam was expanded using a set of extender lenses, yielding a uniform beam profile. An aperture was placed after the extender lens to filter out the poor-quality light at the beam edge. The femtosecond laser was imported into a galvano mirror, which enabled two-dimensional plane scanning. Therefore, three-dimensional movement was realized by combining with the Z-axis piezoelectric platform and motorized stage. After passing through the galvano mirror, the femtosecond laser spot was projected onto an objective (Zeiss, 63X, Oil, NA = 1.4) using a 4F lens system. A dielectric mirror with only 800 nm reflection was used to reflect the femtosecond laser and transmit the illumination light to a Charge Coupled Device monitor. By designing the required structural model, the data coordinates of the structure were directly imported into the fabrication software. The optical performance of the fabricated hydrogel optical elements was measured using a custom imaging setup. In the measurement, the pH values solution is prepared in advance and controlled by the pH meter. Finally, the tunable focusing performance can be obtained by injecting the different pH value solutions into a customized transparent glass box housing the sample.

 

Fig. 1. Schematic experimental setup of the FsLT system.

Download Full Size | PDF

3. Results and discussion

The lithography threshold and surface quality (morphology and roughness) of micro-optical elements critically determine the optical performance of hydrogel materials, implying that the FsLT-induced photo-crosslinking effect should be explored in detail. In a previous study [3], the roughness of the BSA microstructures induced by a femtosecond laser was in the order of hundreds of nanometers, making them unsuitable for optical applications. Thus, fabrication parameters are important for hydrogel microstructures, with several studies reporting respective femtosecond laser-based fabrication systems and hydrogel solutions for Ti/Al₂O₃ solid-state femtosecond lasers [14,17]. To obtain suitable BSA hydrogel structures, the fabrication results of the selected laser-processing parameters for the femtosecond fiber laser are shown in Fig. 2. Figures 2(a) and 2(b) show the optical and fluorescence microscopy images of BSA hydrogel squares obtained for average laser powers of 25 mW, 30 mW, 35 mW, and 40 mW (from left to right) and scanning speeds of 100 µm/s, 200 µm/s, 300 µm/s, and 400 µm/s (from top to bottom). Figure 2(c) displays the color map for the fabrication quality obtained from Figs. 2(a). These results suggest that the lithography threshold for hydrogel polymerization was determined by the scanning speed and laser power. As the scanning speed increased, a higher laser power was required for lithography generation. Based on the lithography morphology and surface quality results, the fabrication parameters for the laser power of 30 mW and scanning speed of 100 µm/s were adopted in our experiments. We also studied the scanning layer interval in the 100–300 nm range (the adjacent layer distance in Z-direction), and obtained the lithographic quality of the FZP, as shown in Figs. 2(c–e). When the layer interval is below 100 nm, the repetitive fabrication area can be easily damaged if the laser power is changed by the laser or environment change or there exist mite defects, for instance, bubble or impurity et al. Evidently, the layer interval of 100 nm yielded the best morphology and surface quality. When the layer interval is larger than 300 nm, the surface quality become poor as the layer interval increasing. Moreover, it is difficult to induce polymerization and repeat the fabrication results, the structure will be sometimes stacked together or washed away when developing the fabricated microstructures.

 

Fig. 2. (a–b) Optical and fluorescence microscopy images of the topography of protein squares (10 $\mathrm{\mu}\textrm{m}\; $× 10 $\mathrm{\mu}\textrm{m}\; $× 5 $\mathrm{\mu}\textrm{m}$) fabricated with different average laser power (range, 25–40 mW) and scanning speed (range, 100–400 $\mathrm{\mu}\textrm{m}$/s), with the BSA concentration of 500 mg/mL and the Rb concentration of 2 mg/mL. (c) The color map for the fabrication quality obtained from (a) and (b). (d-e) Optical microscopy images of the topography of the Fresnel zone plate fabricated using a laser power of 30 mW, scanning speed of 100 $\mathrm{\mu}\textrm{m}$/s, and layer interval in the 100–300 nm range. Scale bar:$10\mathrm{\ \mu m}$.

Download Full Size | PDF

After optimizing the laser lithography parameters, we fabricated separate micro-optical elements and confirmed their optical properties and stimulated responsive tunability. Because the microlens and FZP will work at various wavelengths, we design the morphology of the microlens (diameter: 30 µm and height: 5 µm) and the FZP (diameter: 30 µm and height: 3 µm) and obtained the focal lengths at various wavelengths using FDTD simulation. As shown in Fig. 3(a), a 2 × 2 BSA FZP array (diameter, 30 µm; height, 3 µm) was fabricated on a transparent cover glass using a laser power of 30 mW, scanning speed of 100 µm/s, and the dot, line, and layer intervals of 100 nm. It exhibited acceptable optical surface quality for soft biological materials that are usually difficult to structure using conventional fabrication methods. In Figs. 3(b–d), the focusing performance is demonstrated using a custom imaging setup, for different light wavelengths. Evidently, the BSA FZP array exhibits four uniform focal points at the wavelengths of 633 nm, 532 nm, and 405 nm; the corresponding focal lengths are 67 µm, 81 µm, and 92 µm, respectively. These results are consistent with the variation in the theoretical lens focal length. Further finite different time domain (FDTD) simulations were performed to support the focal length variation as a function of wavelength, as shown in Figs. 3(e–g); the refractive index in the simulations was 1.447, according to the literature [14]. By comparing the experimental and simulation results, we concluded that the focal length of the fabricated BSA FZP agreed well with that of the designed FZP; the error was within ten percent, which was attributed to the fabrication tolerance and rough setting of the simulation parameters, such as the refractive index. Similarly, a 2 × 2 BSA microlens array (diameter: 30 µm and height: 3 µm) was fabricated, and the results are shown in Fig. 4. It exhibited clear and uniform focal images that were slightly larger than those of the FZP array. The focal lengths at the wavelengths of 633 nm, 532 nm, and 405 nm were 87 µm, 92 µm, and 94 µm, respectively. Notably, the focal-length variation of the microlens was smaller than that of the FZP, which was also consistent with the physical theory. The FDTD simulation confirmed the theoretical focal length at different wavelengths in Figs. 4(e–g), and the focal length of the fabricated BSA microlens agreed well with that of the designed microlens shown in Fig. 4(h). Therefore, FDTD simulations can be used for designing achromatic RDMOEs.

 

Fig. 3. (a) Optical microscopy images for the topography of the BSA FZP (diameter, 30 $\mathrm{\mu}\textrm{m}$; height, 3 $\mathrm{\mu}\textrm{m}$) fabricated using an average laser power of 30 mW and scanning speed of 1$00\; \mathrm{\mu}\textrm{m}$/s, with BSA concentration of 500 mg/mL and Rb concentration of 2 mg/mL. (b–d) Focusing results obtained by irradiation with 405 nm, 532 nm, and 633 nm light wavelengths. (e–f) FDTD simulation results of the BSA FZP and (h) simulated and experimental results, as a function of the light wavelength. Scale bar: 10 µm.

Download Full Size | PDF

 

Fig. 4. (a) Optical microscopy images for the topography of the BSA microlens (diameter, 30 $\mathrm{\mu}\textrm{m}$; height, 5 $\mathrm{\mu}\textrm{m}$) fabricated using an average laser power of 30 mW and scanning speed of 1$00\; \mathrm{\mu}\textrm{m}$/s, with BSA concentration of 500 mg/mL and Rb concentration of 2 mg/mL. (b–d) Focusing results obtained by irradiation with 405 nm, 532 nm, and 633 nm light wavelengths. (e–f) FDTD simulation results of the BSA microlens and (h) simulated and experimental results as a function of the light wavelength. Scale bar: 10 µm.

Download Full Size | PDF

At the same time, achieving achromatism in a single material requires tunable performance. Therefore, we tuned the focal length by modulating the pH value under different irradiation wavelengths, as shown in Fig. 5. From Fig. 5(a), the focal lengths of the BSA microlens is observed to have changed dynamically with pH; the variation range was 87–114 µm for three wavelengths. Because the shape of the BSA microlens is circle, the shape will swell and shrink while changing pH value. Thereby the curvature change is large. As a result, the dependence of focal length on the pH value is obvious. The smallest focal length was obtained at pH = 7, whereas the maximal focal length was obtained at pH = 13. The chromatic aberration of the BSA microlens reached 5 µm for the three wavelengths. Thus, using single microlens cannot achieve the achromatism, even under the condition of pH changing (∼5 µm). Therefore, the employment of RDMOE is necessary. Besides, the variety of focal lengths show a symmetric behavior. This symmetric characteristics of the focal lengths as pH changing enable this BSA hydrogel microlens working in both acid and alkali environments by adjusting the tiny distance. Furthermore, the achromatic aberration effect can be achieved in acid and alkali environments with one device. In contrast to the BSA microlens, the focal lengths of the BSA FZP exhibited marginal dependence on the pH, as shown in Fig. 5(b), which has also been reported in the literature [4]. For the FZP, the shape is approximately square and the swelling or shrinking is limited in z direction, changing the thickness. The focal length of FZP is determined by $f \approx {r_m}^2/m\lambda $, which means that the radius ${r_m}$ and wavelength λ are two crucial factors in determining the focal length. Changing pH can change the thickness of the FZP but not ${r_m}$, which is limited by the substrates. Thus, the ${r_m}$ of the BSA FZP always remains constant for pH tuning. Thereby the dependence of focal length on the pH value is not obvious for FZP. Based on above discussion, optical focusing performance can be adjusted by varying the pH, which can be used for achieving hydrogel-based achromatic RDMOEs.

 

Fig. 5. Focal length of the BSA hydrogel (a) microlens and (b) FZP as a function of pH, for the light wavelengths of 405 nm, 532 nm, and 633 nm.

Download Full Size | PDF

Based on the simulation results for a single BSA microlens and FZP, we designed a hydrogel-based achromatic RDMOE. The design principles and processes are shown in Fig. 6. Based on the BSA microlens, we designed the basic structures, including the microlens radius, curvature, FZP, refractive index n, and wavelength, and obtained the initial focal length. Next, we varied the pH by varying the microlens thickness and curvature, hydrogel refractive index n, and FZP thickness, obtaining the simulation results of the RDMOE for one wavelength. Subsequently, we adjusted the other two wavelengths, and mapped the focal length dependence of this RDMOE on pH for the three wavelengths. Finally, by contrasting the focal lengths at the three wavelengths, achromatic aberration results were obtained. Note that the desired achromatic aberration result could be obtained by resetting the basic parameters and repeating the design process. The original and adjusted parameters of the RDMOE are listed in Table 1. The original parameters were obtained for pH = 7, with the microlens radius and curvature at 30 µm and 34 µm, respectively, and the FZP thickness at 1 µm. The refractive index n of the BSA hydrogel and the shrinkage ratio were obtained from the literature [14]. Based on the data in Table 1, the model parameters for the RDMOE at different pH values were obtained; the actual models are shown in Fig. 7. The FZP is at the bottom, whereas the microlens is at the top. Figures 7(a) and 7(b) show the top and side views of the RDMOE; the RDMOE models are shown in Figs. 7(c–f) at pH = 1, 5, 7, and 13, respectively. These models can be further converted into the fabrication data type for fabricating hydrogel-based RDMOEs. Note that the top microlens and the bottom FZP are fabricated as an entirety and there is no alignment in the fabrication of the RDMOE.

 

Fig. 6. Design of the hydrogel microlens-based and FZP-based RDMOE.

Download Full Size | PDF

 

Fig. 7. Designed RDMOE in FDTD simulations. (a) Top and (b) side views and of the RDMOE (top: microlens; bottom: FZP). (c–f) RDMOE models at pH = 1, 5, 7, and 13, respectively.

Download Full Size | PDF

In Fig. 8, the simulated achromatic aberration results for the BSA RDMOE are displayed, and an achromatic simulation diagram at the three wavelengths is shown by FDTD simulation [41]. The simulated focal lengths of the RDMOE for different pH values are shown in Fig. 8(d), which shows that there exists an approximate focal length for the three wavelengths. In Figs. 8(a–c), the simulation results for the wavelength of 405 nm and pH = 1, wavelength of 532 nm and pH = 7, and wavelength of 633 nm and pH = 5 are shown. The corresponding focal lengths are 75.495 µm, 76.432 µm, and 75.766 µm, respectively (all within ±1 µm), and can be considered as the achromatic aberration. Hence, from the theoretical viewpoint, achromatic aberration can be achieved for RDMOEs and tunable BSA hydrogel-based materials.

 

Fig. 8. FDTD simulation results of the BSA RDMOE with achromatic aberration, for (a) the wavelength of 405 nm and pH = 1, (b) the wavelength of 532 nm and pH = 7, and (c) the wavelength of 633 nm and pH = 5. (d) Simulated focal lengths vs. the light wavelength and the pH value.

Download Full Size | PDF

Table 1. Refractive index n, shrinkage ratio, microlens curvature, and FZP thickness, at different pH values [14].

View Table

Subsequently, we experimentally fabricated a BSA RDMOE using femtosecond laser nonlinear lithography and measured its focusing performance; the results are shown in Fig. 9. The optical microscopy images for white light, 405 nm, 532 nm, and 633 nm irradiation wavelengths of the BSA RDMOE, shown in Figs. 9(a–d), were taken in water because the BSA RDMOE easily retains its morphology in aqueous environments. The reason for choosing these three wavelengths that is because these wavelengths are commonly used as the laser emitting wavelengths, which is used in the laser scanning confocal microscope. Considering the biological optical devices represented by the laser scanning confocal microscope, these three wavelengths are used for the measurement of the hydrogel RDMOE in view of potential application in biophotonics. Moreover, these measurement conditions are relevant for applications in biological photonics because biological environments are often solution-based. Furthermore, we measured the focusing performance and focal lengths for the wavelengths of 405 nm, 532 nm, and 633 nm, at different pH values, and the results are shown in Fig. 9(e). When the pH value increased from 1 to 13, the focal lengths at wavelengths of 405 nm, 532 nm, and 633 nm all showed an increasing tendency followed by a decreasing tendency. The maximal focal length was observed at pH = 1, whereas the minimal focal length was observed at pH = 7. The minimal unit of the focal length was one micrometer (limitation of our custom measurement setup). The curves of the focal lengths for the three wavelengths as a function of pH are shown in Fig. 10, demonstrating achromatic aberration. The curves reveal that there exists the same focal length, for instance, at pH = 1 for the wavelength of 633 nm, at pH = 5 or 13 for the wavelength of 532 nm, and at pH = 7 for the wavelength of 405 nm; the focal length was estimated as 70 µm and the corresponding focusing images are shown in the illustration. In other words, the focusing lengths for the three wavelengths can be limited to below 1 µm, which can be regarded as achromatic aberration. Although the experimental achromatic aberration results agree well with the theoretical values, the focal lengths with different pH values are different with the simulation. It can be contributed to the following two reasons: the first reason is the simulated parameters, in especial the refractive index and microlens curvature, are obtained by the semi-quantitative method in the Ref. [13]. This leads to the deviation between the experiments and simulation. The second reason is that the fabrication tolerance for the BSA RDMOE by FsLT is unavoidable because of the water-soluble fabrication environment and “soft and biological” material property. Therefore, we will research the quantitative refractive index change of BSA structures for the different pH values and improve the fabrication accuracy in the further. So we can qualitatively obtain the achromatic aberration result between the experiments and simulation. However, compared with a single BSA microlens, our fabricated BSA RDMOE can still significantly improve achromaticity. Based on the above discussions, we concluded that achromatic aberration within 1 µm was successfully achieved in a single biological material using the currently described RDMOE and stimulated response for three wavelengths. Consequently, such hybrid micro-optical elements can be used in biomedical imaging and optical biology sensing applications.

 

Fig. 9. Optical microscopy images for the following light wavelengths: (a) white, (b) 405 nm, (c) 532 nm, and (d) 633 nm in water. (e) Focusing results obtained after irradiating by light at the wavelengths of 405 nm, 532 nm, and 633 nm, for different pH values.

Download Full Size | PDF

 

Fig. 10. Experimental results for the BSA RDMOE after irradiating by light at the wavelengths of 405 nm, 532 nm, and 633 nm, vs. pH. The insets are the corresponding focusing images.

Download Full Size | PDF

4. Conclusions

In summary, a novel biological hydrogel-based RDMOE was successfully fabricated to achieve achromatic aberrations using FsLT. First, we experimentally fabricated a single BSA microlens and FZP, and demonstrated stimulated responsive tunability by varying the pH value. Meanwhile, an FDTD simulation was used for validating the experimental results, following which the designed parameters were used as the RDMOE parameters. The designed hydrogel-based RDMOE demonstrated achromatic aberration for three wavelengths, using tunable focusing ability, while the experimental hydrogel-based RDMOE also showed the achromatic aberration for three wavelengths by varying the pH value, obtaining the focal length deviation within 1 µm. The achromatic aberration property of biological micro-optical elements in single hydrogel materials is promising for biomedical imaging and optical biology sensing.