Chemical pattern-directed assembly of IL on surface
The carboxylic acid-terminated chemical patterns (partially degraded octadecyltrichlorosilane, OTSpd) were fabricated on the self-assembled monolayer of octadecyltrichlorosilane (OTS) film using the scanning probe deep oxidation lithography method [9
]. The OTSpd pattern is a high energy, lyophilic surface whereas the OTS background is a methyl-terminated, low energy, lyophobic surface. Based on the wetting-driven assembly approach [10
], liquid can be assembled on the chemical patterns due to the contrast in surface energy [11
]. shows a representative OTSpd disc array. shows the same region after a liquid [Bmim]Cl drop rolled over the OTSpd discs. By comparing with , we found that the IL micro-drops were selectively deposited on the high-surface energy OTSpd chemical patterns. is the optical image of the IL drop arrays assembled on OTSpd patterns. The background is the OTS film. Each dark spot in is an IL drop assembled on the OTSpd disc. The regions shown in and 1b are highlighted within the red box in .
Figure 1 [Bmim]Cl assembles on the OTSpd pattern. a) OTSpd discs fabricated by scanning probe deep oxidation lithography on OTS film. Topography image. The center-to-center distance between two neighboring discs is ~7 μm. b) The same area after [Bmim]Cl (more ...)
reveals that the amounts of IL assembled on each OTSpd disc are similar but not identical. Some discs appear darker than the rest, indicating that more IL was assembled on that OTSpd disc. Correspondingly, in the atomic force microscopy (AFM) topography image (), the height of IL drops varies between 250–800 nm. The AFM image reveals more details about the shape of the IL droplet assembled on the OTSpd discs. A representative high-resolution AFM image of the IL drop is shown in Supporting Information File 1
, Figure S1. The AFM topography image shows that the IL is not a hemispherical drop that covers the whole OTSpd disc. Instead, the IL adopts a Mexican hat shape – a partial drop sitting on top of a precursor layer. In the OTSpd disc center is the partial drop, which only covers the central part of the OTSpd disc and is typically 250–800 nm high. The central drop is surrounded by a rim, which extends out and covers the remaining part of the OTSpd disc. The rim is thickest at the foot of the central drop and gradually becomes thinner as it extends out. The Mexican hat shape indicates that the IL drop co-exists with an IL precursor layer (the “rim”). The observed Mexican hat shape for an IL drop is not a surprise. Since first discovered by Hardy in 1919 [13
], the existence of the precursor layer of a drop on a solid surface has been extensively studied. In fact, the Mexican hat shape has been confirmed as the real shape for most liquid drops on solid surfaces, provided that the drop can be resolved with a sufficient resolution [14
[Bmim]Cl is miscible with water. When the sample shown in was immersed in water, all IL micro-droplets on the patterned area disappeared instantaneously, indicating that the IL micro-droplets were dissolved in water.
Silane-coated IL capsules
Silane molecules react with water to form silanols, which subsequently cross-link with each other using the Si–O–Si covalent bonds and form a polymer network [15
]. Such a silane network is mechanically stable and chemically inert. When silane molecules react with hydrophilic surfaces, a self-assembled silane layer is formed on the surface. The cross-linked silane film can be formed on the IL drop surface as well because water adsorbed there. In our experiment, we incubated IL drop arrays in OTS vapor. We found that an OTS layer covered the IL drop surface, forming a capsule that encapsulated the IL inside (). The capsule had the same Mexican hat shape of the IL assembled on the OTSpd disc. The hemispherical cap shaped drop is in the center on the OTSpd disc and co-exists with the surrounding precursor layer (the rim of the Mexican hat), which covers the remaining OTSpd disc. The height cross-sectional profile along the cyan line in the AC mode topography image () is plotted in (black line), which reveals that the drop in is 300 nm in height. is an optical image of the OTS-coated [Bmim]Cl drop array. The image was acquired under water after 1 h of incubation. Under the optical microscope, interfaces separating the drop and the water can be clearly observed. In contrast, in the control experiment for those IL drops assembled on OTSpd disc without OTS coating after water was applied to the patterned area, the un-coated IL drop was instantaneously miscible with water, and thereby disappeared.
Figure 2 a) A representative OTS-coated [Bmim]Cl drop on the OTSpd pattern. AC mode topography image. b) Optical image of OTS-coated [Bmim]Cl drops on the OTSpd disc array. The imaged was acquired under water. c) The phase image corresponding to the topography (more ...)
Because the coated-drops still existed after incubation, we conclude that the coating was a complete layer which can separate [Bmim]Cl inside the capsule and the water outside the capsule. In the phase image () corresponding to the topography image shown in , the phase signals over the drop and the OTS background are the same. This can be further quantitatively illustrated by the phase cross-sectional profile along the cyan line in which is plotted as the blue line in . In the plot, the phase signal difference between the drop surface (central white region in ) and the OTS film (pink regions in ) is smaller than the noise level. Thus, we conclude that OTS and the drop surface have the same phase signal. The phase signal acquired during the same scanning line and under the same instrumental set-ups represents the surface identity. Since the OTS background is methyl-terminated, we conclude that the vapor treated [Bmim]Cl drop is also covered with a layer of methyl-terminated OTS silane.
During the AC mode imaging, we also varied the tapping amplitude set point in order to examine how the encapsulated IL responds to different external tapping intensities. At a high amplitude set point (99.5% of the free oscillation amplitude), the tip tapped the OTS-coated drop gently. A smooth topography profile of the drop was acquired. In contrast, at a low set point (95% of the free oscillation amplitude), the tip tapped the OTS-coated drop hard, with a high force. Phase signal oscillations were observed when the tip scanned over the drop, as shown in the Supporting Information File 1
, Figure S1. The oscillation at a low set point indicates that the drop was disturbed when the tip tapped it hard which caused the IL inside to oscillate. Hence, the IL inside the capsule was still fluidic. In comparison, under the same low set point, the AFM scan lines over the OTS film background did not show any oscillation because the OTS film was in solid phase. Hence, this control reveals that the oscillation we observed over the drop is the true physical oscillation of the IL inside the drop rather than the electronic oscillation originated from the AFM feedback loop. Therefore, from this experiment we conclude that the coated silane layer only formed at the surface of the IL drop.
Reaction of the OTS-coated IL capsules
Pinholes widely exist in the silane film that was prepared without stabilization [17
]. When the “unstablized” OTS film was imaged using a MikroMasch ultra-sharp AFM tip (~1 nm in tip diameter), no pinholes could be resolved. On the other hand, when the unstablized OTS film is incubated in 11-mercaptoundecyltrimethoxysilane toluene solution, the 11-mercaptoundecyltrimethoxysilane molecules can fill the pinholes in the OTS film, leaving the terminal –SH groups on top. The –SH group can subsequently bond to gold nanoparticles and immobilize them on the surface. Hence, we infer that the size of the pinhole would be around 0.5–1 nm. These pinholes provide spaces for reactions and encounters between the materials encapsulated inside the drop and reactants in the external solvent.
shows OTS-coated [Bmim]Cl drop arrays. The IL inside drops contained 30% (w/w) of FeCl3
. The sample was incubated under 30% H2
solution at 25 °C. FeCl3
is a homogenous catalyst for the decomposition reaction of H2
]. When FeCl3
was added in H2
solution, the H2
decomposed and oxygen bubbles were generated in the solution as the decomposition product. In our experimental set-up, the FeCl3
was dissolved in IL solution, which was encapsulated by the OTS coating and existed as the immobilized capsules arrays on the designated places on the sample surface. We applied one drop (30 μL) 30% H2
solution onto the surface to cover the OTS-coated FeCl3
/IL arrays. The Fe3+
inside the IL was slowly released from the pinholes on the OTS film. The released Fe3+
catalyzed the decomposition reaction of H2
, which generated O2
bubbles. The reaction was monitored by the optical microscope in real time. As shows, after immersion, oxygen bubbles were observed on the patterned area immediately (within 0.5 s after the H2
drop was applied onto the sample). Under the optical microscope, the smallest bubble that can be resolved is around 600 nm in size, which is at the resolution limit of our microscope. At the beginning, these small bubbles randomly appeared at the surface of the IL capsules. However, nearby oxygen bubbles fused together to form large bubbles. The size of the bubbles increased with time. Upon further growth, the fused bubbles took off from the surface and the patterned area became clean. Then, new bubbles appeared at the interfaces of the OTS-coated IL drops. These new bubbles may not always originate from exactly the same spot in the array as the previous bubbles did. However, the bubbles always started from the OTS-coated IL drops in the arrays. The overall decomposition reaction lasted for ~12 h, until all H2
was consumed. During the reaction, all oxygen bubbles generated from the decomposition were observed to originate from the surface of the IL capsules. This fact suggests that the majority of Fe3+
ions did not diffuse into the solution.
Figure 3 H2O2 decomposition reaction catalyzed by FeCl3. The process was recorded by the optical microscope under 30% H2O2 solution. The FeCl3 [Bmim]Cl solution assembled on two 8 × 8 OTSpd disc arrays. The IL drops were protected by a layer of OTS film (more ...)
The potential leaking of IL into solution was assessed through the following experiment: As the first step, we fabricated one 8 × 8 OTSpd disc array on a 1 × 1 cm2
OTS sample. In total, we fabricated three such samples (sample A, sample B and sample C). Within the array, each OTSpd disc has a diameter (D
) of 3.5 μm. For sample A, we coated the array with IL solution and then coated a layer of OTS silane to encapsulate it. Then, each capsule’s volume can be obtained through the flooding analysis using the AFM topography image. On average, the volume for each capsule was about 8 μm3
. Then, we put one drop (10 μL) of deionized water over the 8 × 8 OTS-coated IL capsule array for 2 h. Next, we transferred the drop to sample B and let this drop cover on the 8 × 8 OTSpd disc array on sample B. This drop evaporated in air within 30 min. If the IL leaked out from the capsule during the previous 2 h long incubation over sample A, the dissolved IL would be transferred to sample B. Since the IL would not evaporate with the water, IL would be deposited on sample B. Furthermore, because this IL ([Bmim]Cl) does not wet the OTS film, as demonstrated in , the deposited IL would be selectively concentrated on the high-energy 8 × 8 OTSpd disc array on sample B. Therefore, we can use AFM to characterize the 8 × 8 OTSpd disc array to reveal how much IL was deposited. Figure S2 in Supporting Information File 1
shows a representative image of one OTSpd disc after the drop evaporated over the OTSpd disc array. From the image (Figure S2), we computed that the volume of the IL deposition on this OTSpd disc was 0.044 μm3
We used AFM to characterize all 64 OTSpd discs in the array and computed the IL deposition volume, which yielded an average IL deposition volume of 0.04 μm3
/disc. AFM scan also revealed that no IL was deposited on the OTS surface. In the control experiment, we put one drop (10 μL) of deionized water over the 8 × 8 OTSpd array for 2 h over sample C, which just had a clean 8 × 8 OTSpd array. After the drop evaporated in air, we characterized the 8 × 8 OTSpd array on sample C. No deposit was found, the OTSpd arrays did not change. Hence, we conclude that the material deposited on OTSpd disc on sample B is the leaked IL. After a 2 h long incubation, only 0.5% (v/v) IL inside OTS-coated capsule was slowly released to water. The OTS coating leads to the slow release of the IL.
The potential leaking of Fe3+ from the OTS-coated IL capsule was also studied. We put one drop (30 μL) 30% H2O2 solution onto the surface to cover the OTS-coated FeCl3/IL arrays to initiate the reaction. After 1 h, while the reaction was still proceeding, we used a pipette to transfer the solution onto another clean OTS-coated wafer surface. At this stage, if a large amount of Fe3+ was released into the bulk solution phase, the Fe3+ would have been transferred onto the clean OTS-coated wafer surface as well. Then, we injected additional 30 μL 30% H2O2 solution into this drop. Since Fe3+ is the catalyst in the decomposition reaction, it will not be consumed. On the contrary, it would continue to catalyze the decomposition reaction. Nevertheless, we did not observe any oxygen bubbles generated within this drop. This fact suggests that the concentration of Fe3+ within this 60 μL drop was just too low. The Fe3+ concentration in the original 30 μL drop was just twice as high as that of in the 60 μL drop. Therefore, the Fe3+ concentration in the original 30 μL drop would be low as well. Our data show that the OTS coating on the IL drop surface effectively suppressed the diffusion of Fe3+ into the external solution.
In a separated control experiment, OTS-coated FeCl3-free IL drops were incubated with H2O2 solution. No oxygen bubbles were generated, indicating that Fe3+ was responsible for the decomposition of H2O2. From these experimental results we conclude that the H2O2 decomposition reaction occurred at the IL–OTS–water interface. The reaction occurred either because the Fe3+ ions diffused out of or the H2O2 molecules diffused into the IL capsules through pinholes in the OTS film.