Encryption/decryption and microtarget capturing by pH-driven Janus microstructures fabricated by the same femtosecond laser printing parameters

As schematically represented in figure 1(a), micropillars were fabricated by focusing an fs laser beam (with a wavelength of 800 nm and pulse duration of 75 fs) into a pH-sensitive hydrogel, using an objective lens (Olympus, 50×, NA = 0.8). The sample was mounted on a 3D nanotranslation stage (PI, E727) to determine the pillars' locations and heights. When the distance (0.15 μm) between two micropillars was less than the diameter of a single round micropillar, the two micropillars became one micropillar with an elliptical cross-section. Therefore, we can process micropillars by exposing them twice in adjacent positions, termed dual-scanning. In our previous works [33, 35, 36], we found that the expansion ratio of the hydrogel is directly related to the laser exposure dosage for photopolymerization. In figure 1(a), the black side of the pillar was printed first and the grey side was printed second with the same processing parameters. Due to the laser beam scattering effect of the first printed structure, the second printed structure (grey side) was less polymerized than the first printed structure (black side), leading to the spontaneous formation of heterogeneity in micropillar.

Zoom In Zoom Out Reset image size

Photo-crosslinked hydrogel micropillars contain polymeric backbones with ionic pendant groups, which ionize and develop fixed charges in the polymer network, generating the electrostatic repulsive forces responsible for pH-dependent swelling or deswelling [33, 34, 36, 37]. Small changes around the critical pH can significantly change the mesh size of the polymeric networks. The pendant groups of the anionic hydrogels used in this work were un-ionized below and ionized above the pH threshold (pH = 9) of the polymeric network. As a result, the hydrogel swelled when pH > 9 because of the high osmotic swelling force of ions. The volume change property of the structure is closely related to the laser processing dosage. The lower laser power results in weaker extent of polymerization which will lead to lower resistance to swelling force caused by the change of pH values. Therefore, the microstructures obtained with lower laser power have larger expansion. On the contrary, higher laser power brings in lower expansion ratio of polymerized structures. Therefore, Janus micropillars with two sides with different expansion ratios can be readily obtained with our simple dual-scanning strategy. Since a circular micropillar was processed after the first exposure, the light scattering effect of the existing structure reduced the laser power during the second exposure. Therefore, even if the same laser power is used for the dual-scanning process, the power used in the second processing is relatively lower than the first. Thus, the Janus micropillar bends in the direction of the part obtained in the first exposure. Our dual-scanning fabrication is much easier than traditional fabrication methods because the same laser parameters are used throughout the entire process. It is noteworthy that the pH value of the aqueous environment around the micropillars during the fabrication process is less than 9. The contraction effect in an acidic environment is not significant. Therefore, the micropillar recovers to its original straight position when pH < 9. As shown in figure 1(b), the Janus micropillars with two different colors were fabricated by dual-scanning. The black part represents the first exposure area and the gray part represents the second. According to the above analysis, the micropillars are in a straight state when pH < 9, as the unswollen hydrogels on two sides have the same volume. When the pH value exceeds 9, the hydrogel in the second exposure (gray) area expands more than that in the first exposure area (black), so the micropillar bends in the direction of the part obtained by the first exposure. Figure 1(c) shows the optical micrographs of 'two-pillars' structures in straight and bent states. The inset of figure 1(c) shows the pattern, in which the black semicircles represent the position of the first exposure, and the gray semicircles represent the position of the second exposure. The arrows represent the expected bending direction of the micropillar when the solution's pH exceeds 9. Scanning electron microscope (SEM) images of single exposed micropillars and double exposed Janus micropillars are shown in figures 1(d) and (e). The geometry of the micropillars obtained by our dual-scanning closely resemble the geometries of micropillars obtained by a single exposure. The dual-scanned micropillars have slightly larger diameters. The Janus micropillars remained stable and did not show signs on degradation during multiple cycles of bending and recovering (figure 1(f)). The process of changing shape takes about only 0.2 s (SI video 1 (available online at stacks.iop.org/IJEM/3/025001/mmedia)) and the maximum bending angle (θ, the site at which we measure the angle is shown in the inset of figure 1(b)) is about 31°.

The length of the bent micropillar, in the top view, increases with its growth height until it reaches 7 μm (figure 2(a)). When the height exceeds 7 μm, the length of micropillars in the top view remains stable, indicating that bending only occurs at the top of the micropillars. We also investigated the quantitative relationship between the bending angle of micropillar and the distance between the dual-scanning positions (figure 2(b)). The bending angle reaches its peak when the distance between the scanning positions is about 150 nm. The bending angle gradually decreases as the distance increases above 150 nm.

Zoom In Zoom Out Reset image size

Owing to the high flexibility of the laser printing technique, more complex patterns can be achieved based on the dual-scanning approach. By controlling the processing position and sequence of the two exposures, the number of pillars, spatial location, and bending direction can be easily controlled. A variety of ordered patterns can be constructed as a block of Janus micropillars, which can bend and return to a straight state by tuning the pH value. Figures 2(c) and (g) show a series of patterns with multielement cells. The Janus micropillars were in an initially straight state when the pH value was less than 9. When we raised the pH value above 9, all of the micropillars bent in the same direction, which was determined by the sequence of the dual-scanning processes. This pH-driven shape transformation can be repeated many times, exhibiting significant invertibility and stability. The optical micrographs of the Janus micropillar arrays with three- (figure 2(d)) and four-pillar cells (figure 2(e)) show fine replication results and high uniformity of the dual-scanning process. Six-pillar cell arrays with hexagonal arrangement and eight-pillar cell arrays with octagon arrangement were also designed (figures 2(f) and (g)).

Complex petal-like structures can also be fabricated by deliberately setting micropillars with specific bending directions in a pattern. Figure 3(a) shows the conceptional designs by intentionally varying the local position and the bending directions of Janus micropillars in a large cell to create desired hierarchical structures. Figure 3(b) (pH < 9) and (c) (pH > 9) show that complex petal-like structures can be realized on the basis of the design. We can see that the Janus micropillars bend and assemble in the expected directions to form the desired patterns. Moreover, we also fabricated and tested microactuator arrays formed from Janus pillars in channels (figures 3(d) and (g)), which highlights the flexibility of this fabrication method. The red and blue arrows indicate the diffusion direction of pH > 9 and pH < 9 solutions. As the pH > 9 solution diffuses in the channel, all of the microactuators in the array changed from upright to curved in 1 s. On the contrary, returned to upright positions in 3 s as the pH < 9 solution diffused through the array. The differences between the recovery times of the array were dominated by the pH changes caused by the diffusion of the solution, rather than the responding time of the individual Janus micropillar. (More details can be seen in SI Video 3.)

Zoom In Zoom Out Reset image size

Here we explore the possible applications of the pH-driven Janus microactuator arrays in information encryption and decryption. As mentioned above, the bending direction of the Janus microactuator can be controlled arbitrarily by controlling the dual-scanning parameters. We can utilize microactuators with different bending directions to display particular information, similar to nixie tube, which can present complex information with simple graphics and patterns. Moreover, the Janus microactuators straighten when pH < 9, remaining the same as the ordinary micropillars fabricated by a single exposure process. Based on this phenomenon, information can be encrypted and decrypted by changing the pH value. Figure 4(a) shows the design approach of this information encryption/decryption method. Circles with single black color represent the ordinary micropillars fabricated by single exposure, and dichromatic circles represent the Janus microactuators. The arrows indicate the expected bending directions of these Janus microactuators. Information can be displayed when the pH of solution increases (the right of figure 4(a)). The encryption and decryption of the Arabic numerals '1234' at different pH values is demonstrated in figure 4(b). Furthermore, the encryption and decryption of the letters 'CUHK' (figure 4(c)) and 'USTC' (figure 4(d)) were also realized.

Zoom In Zoom Out Reset image size

The resultant Janus microactuators could be applied in microobject trapping, which is in high demanded in the fields of chemical analysis [38] and biomedical devices [39]. Although many interesting works have been published on microobject trapping, simple and tunable microgrippers are still in demand [16, 17, 40, 41]. The trapping process is sketched in figure 5(a), the solvent containing microobjects is dropped on the sample for trapping particles (figure 5(a I)). The claw-like structures containing four micropillars provide an open space that the microobject can enter with ease when pH < 9 (figure 5(a II)). When pH > 9, these Janus microactuators bend and the claw-like structure grasps the microspheres tightly (figure 5(a III)), which can withstand a certain degree of shaking and external force. Figures 5(b) and (d) show that a PS (polystyrene) microsphere with a diameter of 10 μm can enter the open space with a pitch of 11 μm and escape if it is subjected to upward force. When we raise the pH value of the aqueous solution above 9, the microsphere is caught tightly by the four bent microactuators (figure 5(f)). Flow forces were exerted on the microobject in different directions to test the stability of the trapping. The microsphere remained tightly trapped under upward (figure 5(g)) and leftward (figure 5(h)) forces.

Zoom In Zoom Out Reset image size

Firstly, 0.8 ml of acrylic acid (99%), 1.6 g of N-isopropylacrylamide (98%), and 0.15 g of polyvinylpyrrolid-one (average Mw ∼1 300 000) were combined in 1 ml of ethyl lactate (98%), and then stirred to dissolve completely. Then, 2.5 ml of the solution was mixed with 0.5 ml of dipentaerythritol hexaacrylate (98%), 0.5 ml of triethanolamine (99%), and 100 μl of 4,4'bis(diethylamino) benzophenone/N,N-dimethylformamide solution (20 wt%). The solution was stirred overnight to ensure each component was evenly mixed. The prepared hydrogel precursor was kept in yellow light conditions to avoid unnecessary light exposure.

A Ti:sapphire fs laser system (Chameleon Vision-S, Coherent Inc., USA) was used for direct laser writing. The fs-laser had a central wavelength of 800 nm, pulse width of 75 fs, and repetition rate of 80 MHz, respectively. A 50× objective (Olympus, NA = 0.8) was used to focus the laser beam inside the hydrogel. The laser power was 45 mW and the exposure time of single point was 70 ms, while the focus was an ellipsoid with the size ∼1.2 μm × 0.5 μm × 0.5 μm. The exposure threshold of the material was ∼30 mW. Improper energy could cause unsuccessful results. Over-exposure can occur if the power is too high, resulting in disordered structures, while exposure with insufficient power will make the expansion difference between opposite sides of pillar too small to trigger the pH response. A nanotranslation stage (PI, E545) was used to accurately determine the relative position of the sample and focus. The step pitch (x, y, z) of stage was set at 0.15 μm, 0.15 μm, and 0.4 μm. Each vertical scanning process follows a point-to-point scanning manner. The distance between adjacent points was 0.4 μm, and the exposure time of each point was 70 ms. The stage translation time between adjacent points was less than 1 ms. Therefore, the scanning velocity in the vertical direction can be calculated as ∼0.4 μm/70 ms = 5.7 μm s⁻¹. After being polymerized by the fs-laser, the sample was developed in isopropanol for 15 min to remove the unpolymerized part. A sodium hydroxide solution and diluted hydrochloric acid were dropped on the sample to change the pH value of the environment. The microchannel was prepared manually under a microscope by attaching two glass slides (the depth was ∼170 μm) on a glass substrate. The channel width was adjusted with copper wire (with a diameter of ∼200 μm). The channels were prepared by dropping water on substrate and carefully pressing the slides together. Then, the hydrogel resin was dropped into the channel for laser manufacturing.

The structures were deposited with ∼10 nm of gold and subsequently investigated using SEM (ZEISS EVO18) at an accelerating voltage of 10 keV. The optical images were obtained from a Leica fluorescence microscope (LEICA DMI3000 B). In the microfluidic experiment, the change of pH was realized by the spontaneous diffusion of aqueous solutions with different pH values. It should be noted that, because the hydrogel is very soft, microactuators may be washed down by solution flow or touch each other too tightly. A few actuators in the array may not recovery their upright positions in the pH cycles.

Commercialized PS particles (Tianjin Saierqun Technology Co. Ltd, China) with a diameter of 10 μm were mixed in water at a concentration of about 10–4 g ml⁻¹. For the particle-trapping experiment, the solvent containing microspheres was dropped on the sample. Directional forces were generated by dropping water at different positions around the sample.