In this work, we employ tert-butyl acrylate (t-BA) as the oil phase to create an oil-in-water (o/w) Pickering emulsion stabilized by SiO
2-PEI NPs. By further compressing the t-BA/water interface in the L-B trough, we achieve an enhanced control over particle interfacial assembly by using the L-B and Pickering emulsion polymerization methods that materialize into an asymmetrically nanostructured 2D thin poly(t-butyl acrylate) (P(t-BA)) film. We further demonstrate that when silica nanoparticles are functionalized with interfacial ligands, such as polyethyleneimine brushes, these 2D thin nanostructured films could be deployed in various applications, for example, metal-ion extraction and recovery from water. These films can be further re-enforced with a polyvinyl alcohol/glycerol film that provides excellent sturdiness and flexibility. In addition to the creation of metal ion adsorbent the combined L-B and PEmPTech can also be applied to producing 2D nanophotonic monolayer crystals [
2,
3,
24], mitigating the fragility of the resulting monolayers due to reduced inter-particle cohesion that disrupts the crystal-like ordering of the monolayers. In this work, we suggest this possibility by producing 2D monolayer crystals by spreading a Pickering emulsion of a monomer on the surface of water in an L-B trough. Due to a complex mechanism of surface re-arrangement of the nanoparticles at the oil–water interfaces, a continuous monolayer of silica nanoparticles is produced, which upon polymerization evolves into a 2D nanostructured P(t-BA) film with a highly organized self-assembled monolayer of silica nanoparticles that exhibits iridescence and a hint of structural colors, which are, as already mentioned above, most commonly found in nature and biomimetic materials [
22,
23].
3.1. Synthesis SiO2-PEI NPs
The silica nanoparticles were synthesized according to procedures we have previously reported [
18]. The average diameter of the obtained SiO
2-NPs was 737 ± 4 nm. The SEM image of these NPs is presented in
Figure 1A,B. Further, the SiO
2-NPs were functionalized with TESPN to generate nanoparticles with nitril groups on their surface (SiO
2-CN NPs). Further, the nitril groups were transformed by hydrolysis in acidic conditions in the carboxyl groups, so we have generated carboxyl-bearing silica nanoparticles (SiO
2-COOH NPs). Next, we coupled branched polyethyleneimine b-PEI on the surface of these SiO
2-COOH NPs with the help of the DCC coupling agent to obtain SiO
2-PEI NPs, according to the reaction scheme presented in Figure 3A. The evolution of the zeta potential with the surface chemical reaction is given in
Table 1, and while the starting NPs had a negative surface potential at normal pH, the zeta potential of the SiO
2-CN NPs slightly decreased due to the excess negative charge brought by the nitrile groups. The zeta potential increases with the hydrolysis of the nitrile functional groups and their conversion to carboxyl functional groups, SiO
2-COOH NPs, which are partially protonated at neutral pH = 5.7. SiO
2-PEI NPs have a positive surface potential, which is supporting evidence that the surface functionalization has been successful.
The FTIR spectra of the silica nanoparticles are given in
Figure 2, and
Figure S1 shows the chemical modification following the surface reactions with functionalization agents as follows. Structural vibration modes of the silica nanoparticles are typically observed at 1100 cm
−1 and are associated with asymmetric Si-O-Si vibrations and symmetric Si-O-Si vibrations at 790 cm
−1 as well as Si-O-Si deformation vibrations at 471 cm
−1, which are also typically observed in cyclic siloxanes. Further, also characteristic to silica nanoparticles are the Si-OH stretching vibrations observed at 947 cm
−1 and deformation vibrations at 1634 cm
−1. The broad bands observed around the value of ≈3400 cm
−1 are associated with water bound by hydrogen bridges and the stretching vibrations of the Si-OH bond isolated or linked by hydrogen bridges [
18,
25].
For the SiO2-COOH NPs, the FTIR spectrum shows the disappearance of the -C≡N band at 2251 cm−1, which was found in the starting SiO2-CN NPs and even remnant in the SiO2-COOH NPs after hydrolysis, signaling that PEI modification led to the disappearance of this band. In addition, the successful chemical modification of the surface of the NPs is confirmed by the disappearance of the characteristic vibrations of the carboxyl groups at 1535 cm−1, 1744 cm−1 and 1817 cm−1. In addition, the peaks at 2872 cm−1 and 2901 cm−1 overlap after modification of the nanoparticles with PEI, and we believe this is due to the strong hydration of the PEI corona and appearance of a very broad and intense hydrogen bond band.
3.2. Synthesis of the P(t-BA)/SiO2-PEI NPs-Microspheres
To synthesize P(t-BA)/SiO
2-PEI NPs-Microspheres, we have first generated a stable Pickering emulsion from the monomer t-BA and water containing SiO
2-PEI NPs that stabilizes the emulsions, according to the scheme presented in
Figure 3B. The Pickering emulsions are stabilized due to the interfacial adsorption of the silica nanoparticles at the oil–water interface with the formation of a self-assembled monolayer. The monolayer of nanoparticles acts like a shield preventing the coalescence of the oil droplets [
18,
26]. Thus, these special structures, the liquid Pickering emulsion droplets covered by a shield of self-assembled monolayer of nanoparticles, are often referred to as “colloidosomes” and will be referred to in the same way in the current work.
The emulsion generated was stable and could be thus polymerized at room temperature (RT) under exposure to UV radiation, and this way, all the monomer droplets/colloidosomes formed in the emulsion are converted into P(t-BA) microspheres, see
Figure 1C. The surface of these microspheres is nanostructured due to the trapping of the self-assembled monolayer of nanoparticles. The SEM images in
Figure 1D reveal a highly organized structure of these nanoparticles that pack in a hexagonal closed-packed (HCP) arrangement on the large surface area of the microspheres, forming a perfectly self-assembled film as compared to a less organized structure in the cast film
Figure 1A,B. The optical microscope images of the polymer microspheres,
Figure 3C, reveal multicolored iridescent beads due to the diffraction of light from the ordered structure of the self-assembled monolayer at the surface of the microspheres, a well-known phenomenon observed in photonic crystals or highly ordered crystalline monolayers of nanoparticles [
27].
Figure 3.
Reaction scheme for the surface coupling of b-PEI on the surface of the SiO2-COOH NPs (A). (B) Schematics showing the process of Pickering emulsion preparation from two immiscible phases, the liquid organic phase that is the monomer and the water phase containing a suspension of silica nanoparticles, followed by its conversion into an o/w emulsion and polymerization. (C) Optical microscope images of the P(t-BA) microspheres, showing iridescence due to diffraction of the light from the ordered self-assembled monolayers of silica nanoparticles at its surface.
Figure 3.
Reaction scheme for the surface coupling of b-PEI on the surface of the SiO2-COOH NPs (A). (B) Schematics showing the process of Pickering emulsion preparation from two immiscible phases, the liquid organic phase that is the monomer and the water phase containing a suspension of silica nanoparticles, followed by its conversion into an o/w emulsion and polymerization. (C) Optical microscope images of the P(t-BA) microspheres, showing iridescence due to diffraction of the light from the ordered self-assembled monolayers of silica nanoparticles at its surface.
3.3. Synthesis of the Langmuir–Blodgett 2D P(t-BA)/SiO2-PEI NP Film
The process of forming Janus 2D P(t-BA)/SiO
2-PEI NP films from the Pickering emulsions in an L-B trough is similar to the one we have previously reported [
26]. Briefly, the Pickering emulsion droplets (colloidosomes) were added onto the surface of a Langmuir–Blodgett trough with a spatula. The emulsion droplets exhibit an initially highly energic motion on the surface of the water until they break, forming a film,
Figure 4A. We presume that the energic motion is due to re-arrangement of the interface due to a competitive wetting process, such that the SiO
2-PEI NPs, being hydrophilic, are wetted by water; at the same time these are also wetted by the monomer, the SiO
2-PEI NPs also remain attached to the monomer droplet surface. As we will see next, the SiO
2-PEI NPs exclusively occupy the monomer/water interface. A similar phenomenon was also observed in the case of silica nanoparticles bearing different functional groups, such as glycidyl, and was observed for other particles bearing surface functional groups with a higher component of the dispersive surface energy [
26]. The monomer droplet/water interfacial tension is stabilized by the presence of the nanoparticles. Thus, we hypothesize that the energic motion of the monomer droplet is due to Marangoni gradients acting upon the monomer colloidosome, as well as the capillary forces due to initial wetting of the self-assembled SiO
2-PEI NP monolayer at the surface of the colloidosomes by water.
Next, after approximately 1 mL of Pickering emulsion was spread on the water surface, the L-B barriers were compressed such that the available surface area decreased from initially 100% to ca. 20%; these values refer to the absolute area of the trough and were displayed on the graph of the isotherm during compression. When the barriers of the trough were fully open, the available water surface area was 100%, and when the barriers were closed, this decreased to approximately 20%. Through trial and error, we adjusted the volume spread to 1 mL to obtain a reasonable match between the area occupied by the nanoparticles and the area of the trough at full compression, as verified via SEM,
Figure 4. After the barriers close, the polymerization begins by exposing the film to a UV lamp for 2 h. After the polymerization, see
Figure 4B, the polymer film is collected from the water surface with a tweezer for further handling and processing or on an aluminum stub for scanning electron microscopy (SEM) investigations. The cartoon in
Figure 4B already suggests that the film thickness fluctuation originates from the boundaries of broken colloidosomes spread on the surface of the water. When the surface is supersaturated with the Pickering emulsions, the colloidosomes will not break, and will only be partially integrated into the film, as shown in the SEM images of
Figure 4C.
After drying, the 2D P(t-BA)/SiO
2-PEI NP film obtained appears to be sufficiently rigid for careful handling with tweezers, as shown in
Figure 5A. Interestingly, the membrane exhibits iridescence, showing rainbow colors due to the Bragg diffraction of light upon illumination of the membrane with the lamp from a smartphone at certain incident angles, see
Figure 5B–D, which resembles the properties of photonic crystals or a monolayer crystal of nanoparticles, also prepared via the Langmuir–Blodgett technique [
2,
24,
27]. The optical microscope images in
Figure 5E,F further show the more intricate details of the 2D P(t-BA)/SiO
2-PEI NP film, with a hint of red structural color. In contrast, while these were prepared by the transfer of a fragile monolayer of nanoparticles onto a solid substrate by dip-coating, which may disrupt their organization, the remarkable advantage in the current case is that the great degree of ordering achieved by L-B technique is preserved due to the polymerization of the monomer. Thus, it is this organization that leads to the iridescent structural colors of the films, with potential applications in various fields. We have thus demonstrated a facile method for the fabrication of such 2D highly organized crystalline monolayers supported on a thin P(t-BA) film, which, to the best of our knowledge, is being presented here for the first time.
The SEM images presented in
Figure 6 reveal an asymmetric Janus-like characteristic of the 2D P(t-BA)/SiO
2-PEI NP film. On one side of the film,
Figure 6A, there are no particles to be observed, and this is the side of the film (the back side) that is not in contact with water but with air, the polymer/air interface. On the other side of the film,
Figure 6B,C, a highly organized monolayer of self-assembled SiO
2-PEI NPs can be observed, and this comes from the polymer/water interface of the film (the front side).
Figure 6B shows a large area of ordering of the monolayer of nanoparticles, with an almost perfect hexagonal-closed crystalline packing.
The SEM images in
Figure 6D shows the cross-section of the 2D film, where the asymmetric Janus-like characteristic can be clearly observed, with the SiO
2-PEI NPs populating only one side of the film (on the polymer/water interface), while the other side is completely void of nanoparticles. In addition, the thickness of the P(t-BA) film ranges between 500 and 1200 nm.
Figure 6D provides a snapshot of the high degree of ordering of the NPs on large areas of the film surface; for example, rows of ordered nanoparticles can be seen running from the edge of the cross-section deep into the visual field of the film for tens and hundreds of microns. Such ordered structures have been produced in our group and reported previously in Pickering emulsions [
18,
28].
One disadvantage of the film is its relative fragility; the thin 2D monolayer breaks very easily if not carefully manipulated. Thus, to provide better support and strengthen the film, we have reinforced it by pouring 2 mL of an aqueous solution of a PVA/glycerine mixture onto approximately an 8 cm
2 area of the side of the film free from nanoparticles. This produces a flexible PVA/P(t-BA)/SiO
2-PEI NP film that is sturdier and can be easily handled for further use in applications; for example, it can be deployed in multiple cycles of adsorption and desorption of metal ions from aqueous solutions. The photograph of the obtained film is presented in
Figure S2.
3.4. Application of Thin 2D P(t-BA)/SiO2-PEI NP Films for Ion Adsorption
The 2D P(t-BA)/SiO
2-PEI NP film was next deployed to test its capacity for adsorption and desorption of metal ions from water samples. Its adsorption capacity was compared with that of the P(t-BA)/SiO
2-PEI NPs-Microsphere, SiO
2-PEI NPs and PVA/P(t-BA)/SiO
2-PEI NP films. The active components of binding and capturing metal ions from water are the SiO
2-PEI NPs through the PEI layer at their surface. According to our previous studies [
29,
30], the PEI layer is rather non-specific and capable of binding a variety of different metal ions, albeit with different affinities. However, the use of nanoparticles in real applications for ion extraction is not desired as they may escape into the surrounding medium, posing a potential environmental threat themselves. However, silica nanoparticles are generally considered neutral to the environment, as they are already largely encountered in the current form in nature [
31]. Thus, in this study, we measure the evolution of the adsorption capacity of the SiO
2-PEI NPs in different states in powder form as self-assembled monolayers on the surface of microspheres in a 2D L-B film and in an imbedded PVA film. The capacity of these materials was tested for adsorption of Cu(II) metal ions from aqueous solutions. As already alluded to, the extraction capacity
qe refers to the extraction of metal ions from a solution containing the metal ion, and in this case, the difference between the concentration of the initial solution and that of the solution after being in contact with the adsorbent for 12 h was measured according to Equation (1). On the other hand,
qr refers to the recovery of metal ions from the material using an acidic solution, and the recovery capacity of metal ions was calculated according to Equation (2).
The calculated adsorption capacities,
qe and
qr, for Cu(II) are provided for each sample in
Figure 7. Several conclusions can be drawn from the presented data. First, the adsorption capacities decrease in the following order: 2D P(t-BA)/SiO
2-PEI NP film (
qe = 8.4 mg/g,
qr = 12.8 mg/g) > SiO
2-PEI NPs (
qe = 5.8 mg/g,
qr = 10.8 mg/g) > PVA/P(t-BA)/SiO
2-PEI NP films (
qe = 5.0 mg/g,
qr = 8.6 mg/g) > P(t-BA)/SiO
2-PEI NPs-Microspheres (
qe = 0.4 mg/g,
qr = 0.9 mg/g). Given that the SiO
2-PEI NPs are the only active component for binding the metal ions, one might expect them to exhibit the highest adsorption capacity. However, the results suggest otherwise. This discrepancy could be partially explained by the close numerical results between 2D P(t-BA)/SiO
2-PEI NP films and SiO
2-PEI NPs, which fall within the experimental error margin of the measurement. The large experimental error, especially in the case of the 2D P(t-BA)/SiO
2-PEI NP film, is due to the difficulty in handling an ultrathin sample film, which was solved by reinforcing it with PVA. The further decrease in the capacity of adsorption, significant for the P(t-BA)/SiO
2-PEI NPs-Microsphere and only slightly for the PVA/P(t-BA)/SiO
2-PEI NPs, can be explained by addition of inert materials, P(t-BA) and PVA, which theoretically absorb little or no Cu(II) ions. Nonetheless, it is notable that the 2D P(t-BA)/SiO
2-PEI NP film and the PVA/P(t-BA)/SiO
2-PEI NP film exhibit a better adsorption capacity than that of nanoparticles and could potentially be used as an adsorbent on water surfaces for environmental remediation [
31], which also poses the risk of losing expensive adsorbent nanomaterials. Therefore, the 2D P(t-BA)/SiO
2-PEI NP films or PVA/P(t-BA)/SiO
2-PEI NP films provide similar adsorption capacities to that of the nanoparticle powder but have an improved sample handling and control. To further emphasize the ease of handling, especially for the PVA/P(t-BA)/SiO
2-PEI NP films, we provide an image of the film in
Figure S2.
The current results prove that the adsorption capacity for Cu(II) metal ions makes these 2D thin nanostructured films extremely competitive to other nano-engineered material adsorbents [
32,
33]. For a literature comparison, Janus nanoparticles modified with PEI brushes have shown an adsorption capacity for Cu(II) ions of 6 mg/g [
29], which aligns well with the capacity of SiO
2-PEI NPs (
qe = 5.8 mg/g,
qr = 10.8 mg/g) in this study. For the wax colloidosomes generated with the same Janus nanoparticles, as reported by Pauli [
29], the adsorption capacity for Cu(II) was ≈0.5 mg/g, which is comparable with that of the P(t-BA)/SiO
2-PEI NPs-Microsphere (
qe = 0.4 mg/g,
qr = 0.9 mg/g) obtained in this work. Conversely, Tan et al. [
30] have reported an adsorption capacity for Cu(II) for polystyrene nanoparticles modified with PEI of between 20 and 40 mg, which is at least double that of the SiO
2-PEI NPs reported here. To our knowledge, there are no similar reports for 2D membranes of adsorbents thus far, highlighting the novelty of the current combined L-B and Pickering emulsion technologies to produce a 2D membrane of nanoparticles capable of metal ion adsorption.
In addition, the fact that the nanoparticle layer is found in a highly organized form on the surface of the P(t-BA) film or microspheres makes it attractive for a combination of applications, such as nanophotonic structural colors [
22], photonic pigments [
1], biomimetics [
23], or sensors. This current method of manufacturing 2D, thin, supported monolayer crystals combining the L-B method and Pickering emulsions can be used without adaptation for other types of nanoparticles, rods, and nanostructures.