Molecularly Imprinted Porous Aromatic Frameworks for Molecular Recognition.

Porous aromatic frameworks (PAFs) are an important class of porous materials that are well-known for their ultralarge surface areas and superb stabilities. Basically, PAF solids are constructed from periodically arranged phenyl fragments connected via C-C bonds (generally), which provide vast accessible surfaces that can be modified with functional groups and intrinsic pathways for rapid mass transfer. Molecular imprinting technology (MIT) is an effective method for producing binding sites with a specific geometry and size that complement a template object. This review focuses on the integration of MIT into PAF structures via state-of-the-art coupling chemistry to expand the application of porous materials in the fields of metal ion extraction (including the nuclear element uranium) and selective catalysis. Additionally, a concise outlook on the rational construction of molecularly imprinted porous aromatic frameworks is discussed in terms of developing next-generation porous materials for broader applications.

Introduction

In biological and chemical processes, molecular recognition is a crucial step that governs the capabilities of enzymes and receptors in biological functions.[1] In principle, the subtle structure (including a tailored configuration and well-designed functional groups) forms specific interactions with the target substrate.[2] Because of this precise combination, organisms achieve efficient catalysis or response processes. As a consequence, the design and manufacture of enzyme and receptor mimics have long been pursued to efficiently identify and convert substrate molecules for chemical production, drug testing, artificial organs, and other applications.

Inspired by nature, a variety of synthetic systems have been developed for molecular recognition including supramolecular amphiphiles, cavity inclusion, and dynamic combinatorial/metallo-capsule/polymer receptors, etc.[3−5] Significantly, molecularly imprinted technology is considered to be an effective and efficient approach for realizing the molecular recognition abilities.[6] Generally, this technology is achieved through the following steps (Figure ): (i) The template (ion, molecule, macromolecular assembly, and microorganisms) and functional groups form an imprinted complex by a self-assembly process involving multiple interactions such as van der Waals forces, hydrogen bonding, π–π interactions, ionic interactions, and coordination bonds. (ii) The imprinted complexes are incorporated into a bulk polymer through cross-linking agents, which facilitates the fixing of the position of the respective group. (iii) After the removal of the template, the final structure contains cavities that are capable of recognizing and rebinding the target objects and their analogues.[7,8] Correspondingly, the generated molecularly imprinted polymers (MIPs) exhibit several leading edges, including high physical stability, specific recognition, a predictable structure, and universal application.[9,10] Thus, molecularly imprinted technology has attracted widespread attention for applications such as chromatographic separations, artificial antibodies, sensing, artificial immunoassays, drug delivery, and catalysis.[11]

Figure 1

Five main types of molecular imprinting: (i) noncovalent, (ii) electrostatic or ionic, (iii) covalent, (iv) semicovalent, and (v) metal coordination. An imprinted complex composed of the target object and ligands with functional groups is formed through several binding patterns: (I) hydrogen bonding, van der Waals, and π–π interactions; (II) electrostatic or ionic interactions; (III) a covalent bond; (IV) a covalent bond with a spacer; and/or (V) ligand–metal coordination. The ligand Y contains a reactive group for the cross-linking reaction. Then, the imprinted complex with the linker molecules is copolymerized to form the polymer matrix (gray). After removing the template object, an imprinted site is left behind with functional groups fixed on the polymer walls. Finally, the imprinted site with the tailored structure and well-designed functional groups rebinds the target objects. Reprinted with permission from ref (10). Copyright 2014, Royal Society of Chemistry.

Despite the tremendous success in molecular recognition, many problems with respect to its performance restrict its wide utilization.[8−10] (1) The microrheology of the polymer distorts the spatial position of the functional groups, which then lose their selective capability to recognize specific template molecules. (2) Due to the dense structure of the polymer originating from the flexible skeleton, few imprinted sites are exposed on the particle surfaces, and a large number of imprinted sites are entrapped in the interior of the grains, thus greatly reducing the utilization of the imprinted sites. (3) Target objects with a relatively large diameter cannot be effectively transported through the channels. Although MIPs reveal significant characteristics for diverse applications, they suffer from some burning issues, including template leakage, a low binding capacity, and a slow diffusion velocity.

Porous materials with nanometer-size pore cavities are regarded to be a hot research topic in the chemistry and material science fields.[12,13] This Outlook aims to reveal the unique advantages of porous materials for the application in the molecular recognition field. Zeolites were the first well-studied member and opened the door for the systematic investigation of composition, structure, properties, and functions.[14] The excellent performances of zeolites in gas adsorption and separation and in catalysts[15] have motivated the rapid development of other porous solids including aluminophosphates (AlPOs), mesoporous materials (OMMS), and metal–organic frameworks (MOFs).[16−18] These porous materials with open architectures have numerous accessible surfaces and large storage spaces, enabling full host–guest interactions and rapid diffusion velocities. Based on these unique physical and chemical characteristics, they not only exhibit excellent properties in gas separations, molecular storage, and catalysis but also can be used in an extensive range of applications, such as monitoring, drug release, photoelectricity, etc.[19−23] For instance, a copper MOF composed of 3,6-di(pyridin-4-yl)-1,2,4,5-tetrazine (dptz) as the ligand shows a negative solvatochromic behavior with a blue-shift of the absorption band with the increase of solvent polarity (Figure ).[24,25]

Figure 2

UV–vis spectra and photograph of an MOF-based sensor containing solvent molecules. Reprinted with permission from ref (24). Copyright 2011 American Chemical Society.

Porous materials with diversified structural units can be endowed with talented architectures (including specific groups and suitable pore/window sizes), which direct the unique roles for guest molecules via several types of interactions: (i) van der Waals interactions with the porous channels, (ii) coordination bonds with metal centers, (iii) hydrogen bonding interactions with the framework surface, and (iv) π–π interactions with phenyl fragments.[26] Two remarkable examples demonstrate the conventional strategies for achieving specific objectives in porous materials (Figure ). (1) The ultra-micropores in zeolites allow us to rationally adjust to arrange multiple small gas molecules exclusively by size, which can be used to achieve gas separation into high purity grades for practical industrial applications.[27,28] (2) MOF samples are gifted with open metal sites that can strongly interact with guest molecules such as hydrogen and acetylene.[29−33] Therefore, they could potentially be capable of separating gas molecules and sensing neutral and ionic species. However, these materials have many drawbacks, such as low selectivity, complicated synthesis procedures, and limited universality.

Figure 3

(a) Zeolites for CO2/CO/O2 separation. Reprinted with permission from ref (27). Copyright 2018, American Chemical Society. (b) Acetylene separation based on metal–organic frameworks with open metal sites. Reprinted with permission from ref (29). Copyright 2019, American Chemical Society.

Inspired by nature, the specific interaction for a target object is realized through the assembly and arrangement of various functional groups in a confined cavity. Scientists aim to mimic this mechanism by decorating the porous scaffold with the functional groups using two strategies—pre- and postsynthesis modification (Figure ).[34−37] Presynthesis modification uses a building block with functional groups; after the cross-linking process, the functional groups are introduced into the porous network. Nevertheless, functional groups sometimes interfere with the reactivity and reduce the degree of the polymerization reaction to some extent. Furthermore, some prereactions (for instance, acid–base reactions and metal–ligand coordination) between the different functional groups lead to the deficiency of their activity.[38−41] In addition, the cooperative effect of multiple functional groups is difficult to achieve when they are randomly distributed in the porous matrix. In postsynthesis modification, the porous framework is constructed first, and then, the architecture is decorated with functional groups.[42−44] Some problems are inevitably encountered in this method: (1) The diffusion of functional groups through a porous solid results in a concentration gradient, and the amount of functional groups is high on the outside of the particle and low inside the particle. (2) The functional groups are randomly distributed, making it difficult to achieve cooperative effects. These problems make it an extravagant hope to fix functional groups in an angled and related pattern for molecular recognition.

Figure 4

Decorating diversified functional groups into the porous skeleton via presynthetic (a) and postsynthetic (b) modification.

As previously mentioned, MIT technology could be used to integrate various functional groups into tailor-made binding sites to complement the geometrical shape and chemical composition of target molecules. However, a molecularly imprinted complex consisting of multiple functional groups and a template object has an asymmetrical structure that does not match the fragment of the porous framework, leading to the distortion of the porous skeleton. After ditching the template molecules, the dynamic skeleton in classical materials (MOFs, for instance) undergoes a self-assembly process. Subsequently, the relative positions of the functional groups in the imprinted sites change, resulting in the deactivation of the imprinted pattern.

Recently, porous organic materials have attracted much attention because they combine the advantages of polymer materials and inorganic porous materials. They have several unique advantages, such as a low density, a large specific surface area, structural diversity, and tailorability. Their geometry can be predefined by the rational design of the size, structure, and connectivity of the building blocks. This control enables the fine regulation of the chemical and physical properties of the porous skeleton, resulting in an ideal environment for gas adsorption/separation and molecular storage, and the accessible and definable surface argues in favor of catalysis and sensing applications.[45−60]

In these cases, constructing diverse structures becomes the main concern for the state-of-the-art applications. With the help of the different preparation methods, porous organic materials connected by covalent bonds can be classified as hyper-cross-linked polymers (HCPs),[61,62] polymers of intrinsic microporosity (PIMs),[63,64] covalent organic frameworks (COFs),[65−67] conjugated microporous polymers (CMPs),[68,69] covalent triazine frameworks (CTFs),[70] porous aromatic frameworks (PAFs),[71] covalent organic polymers,[72] and porous polymers,[73−75] etc.,[76−83] in a timed sequence (Figure ). These materials have various structural characteristics: (1) HCPs are obtained by pillaring solvent-swelled polymers, generally through Friedel–Crafts alkylation, to expand the dense structure of flexible polymers. The pore cavities of PIMs with a 1D polymer skeleton are generated by distorting rigid fragments. COFs are prepared via the reversible cross-linking of polyhedral monomers to form a thermodynamically controlled skeleton and crystalline structure. CMPs have two typical features, microporosity and a π-conjugated structure, as the name implies. CTFs are derived from the cyclotrimerization of nitrile groups to obtain triazine ring segments in porous networks. PAFs are emanated through irreversible coupling reactions (generally C–C bonds) and endowed with an ordered local structure.[84−87]

Figure 5

Important discoveries in porous materials in a timed sequence.

Based on their structural characteristics, PAFs may emerge as materials that have a positive effect on molecular recognition. The strategy of combining MIT technology with PAF frameworks is investigated, and the performances of the resulting materials (MIPAFs) are fully analyzed to guide the future development of porous solids for advanced applications. The emergence of MIPAF materials, which are amorphous, has significantly enriched the field of porous materials.

The first PAF material was derived from the diamond structure with phenyl rings inserted into the C–C interspace to expand the porous architectures.[71] The tetrahedral building block tetrakis(4-bromophenyl)methane was cross-linked to itself via a C–C Ullmann coupling reaction to obtain PAF-1 with a specific surface area of 5600 m2 g–1 based on the Brunauer–Emmett–Teller (BET) model (Figure ). PAF-1 exhibits excellent thermal stability (>520 °C in air) and solvent stability in moist/acidic/basic environments. Due to the ultralarge surface area, it demonstrates high uptake capacities for hydrogen (10.7 wt % at 77 K, 48 bar) and carbon dioxide (1300 mg g–1 at 298 K, 40 bar).

Figure 6

Construction of the PAF-1 sample through the Ullmann coupling reaction. Reprinted with permission from ref (63). Copyright 2009, Wiley-VCH.

Based on this solid foundation, the synthesis of PAF materials has been continually developing by adjusting the structural composition, pore size, and channel environment.[88−94] From a synthesis perspective, the building monomers are usually composed of phenyl fragment-based building units with highly rigid and symmetrical geometries, including linear, trigonal, tetragonal, and tetrahedral structures.[95−98] Regarding the polymerization reactions, a diversity of linking approaches, such as Yamamoto-type Ullmann, Suzuki, Sonogashira–Hagihara, and Heck coupling reactions, are effective for the construction of PAF networks.[99−105] The wide availability of monomers together with the diversity of coupling modes provides convenient conditions for the design and synthesis of various PAF solids.

Incorporating Single-Function Imprinted Sites for the Removal of Heavy Metals

Lead is an essential substance in the battery and paint industries. With the rapid development of industrialization, the global demand for refined lead increases sharply by 1.29 million tons from 2012 to 2016. According to statistics, only 4.9% of total consumption is effectively recycled for commercial usage every year. Correspondingly, a large number of lead ions are discharged into the natural environment, causing serious environmental pollution and lead poisoning. Due to the great pressure of environmental and economic problems, the establishment of an efficient lead recycling system is urgently needed. The authors mixed methacrylic acid (MAA), 4-vinylpyridine (4-VP), and Pb(NO3)2 together to obtain a Pb-imprinted complex (Figure a).[106] The pyridyl group acts as a monodentate ligand to bind to a Pb2+ ion through a coordination bond. The deprotonated carboxyl group also coordinates to the Pb2+ ion, which leads to the cooperative effects of the pyridine ring and carboxyl group in the Pb-imprinted complex. This Pb-imprinted complex contains vinyl groups at the end, which can undergo a Heck coupling reaction to realize a PAF skeleton extending around the imprinted complex. The Pb-imprinted complex was used as a building unit, replacing the original monomer (divinylbenzene) at molar ratios varied from 0%, 10%, 20%, 30%, to 40% to obtain PAF-10a, PAF-10b, PAF-10c, PAF-10d, and PAF-10e, respectively (Figure b). After removing the Pb2+ centers from the PAF structure to obtain MIPAF, the carboxylic acid and pyridine groups are fixed in the PAF framework by rigid structural segments.

Figure 7

(a) Structural units for MIPAF materials 10b–10e. (b) Ion adsorption of PAF-10d against various interfering ions. (c) Comparison of selectivity and saturation uptake for Pb2+ ion. Reprinted with permission from ref (106). Copyright 2018, Royal Society of Chemistry.

These MIPAF materials maintain the large porosity of the parent PAF sample (613 m2 g–1) with surface areas ranging from 87 to 427 m2 g–1. The utilization of the imprinted sites varies from 96% for PAF-10b to 57% for PAF-10e. The largest sorption uptake of 90.36 mg g–1 is achieved with MIPAF-10d (utilization of imprinted sites ∼89%). In contrast, the traditional polymer-based MIP powder with the same amount of imprinted complexes has a specific surface area of 37 m2 g–1. Its capacity for Pb2+ ions is only 5.5 mg g–1, corresponding to the utilization of 14% of the imprinted sites. The capacity of MIPAF-10d is ca. 15 times higher than that of MIP. All these results demonstrate that the large accessible surface in the MIPAF architecture provides a larger number of imprinted sites for Pb2+ ion binding (Figure b). The selectivity coefficient of MIPAF-10d for Pb2+ ions exceeds 1.5 × 102 (Pb2+/Cu2+), which is more than 3 times that of MIP (Figure c). This phenomenon is attributed to the fact that the rigid scaffolds of PAFs effectively keep the relative positions of the functional groups as they were in the imprinted complex, and the two functional groups cooperatively and selectively bind lead ions again.

Because of the pure organic composition of the PAF network, the MIPAF-10d powder can be charged into a CH2Cl2 solution of poly(methyl methacrylate) (PMMA) to prepare a composite pipe using MIPAF-10d as an interior coating. This MIPAF-10d-doped composite pipe selectively adsorbs emitted Pb2+ ions with high efficiency (>99.94%). The Pb2+ ion concentration in the outflow decreases from an original 20 ppm to ∼0.008 ppm, satisfying the standard issued by the U.S. Environmental Protection Agency (0.015 ppm).[90,91]

Incorporating Dual-Function Imprinted Sites for Tandem Functions

Natural enzymes have specific recognition and catalytic properties toward substrate molecules. A variety of enzymes cooperate to achieve efficient metabolism in the human body. Many scientists hope to mimic these interrelated and collaborative catalytic processes to realize an efficient production technology. However, different functional groups might interact with each other to form acid–base or coordination adducts, rendering them ineffective and preventing them from occupying specific fixed positions. Using the MIT method to preassemble the functional groups needed for various target capabilities, different types of imprinted sites may be embedded in the large accessible surface of the PAF platform to perform tandem or cooperative behaviors that originate from the rigid skeleton and large accessible surface.

Organophosphorus compounds are a kind of nerve agent that can seriously damage the nervous system of the human body. In order to degrade this poison, the authors designed a hydrolysis site for organophosphorus that simulates the binuclear zinc centers in the natural enzyme.[107] A series of MIPAFs were synthesized via a Heck coupling reaction with 10%, 20%, 30%, and 40% of hydrolysis site (molar ratio of hydrolysis complex to p-divinylbenzene) and are denoted as MIPAF-1, MIPAF-2, MIPAF-3, and MIPAF-4 (Figure a). The BET surface areas of MIPAF-1, MIPAF-2, MIPAF-3, and MIPAF-4 are 503, 390, 275, and 160 m2 g–1, respectively. For MIPAF-1 through MIPAF-3, the hydrolysis speed gradually increases, and then, it decreases with the increased content of hydrolysis sites in MIPAF-4. The hydrolysis rate of MIPAF-3 is 9.5 × 10–6 m min–1, which is ∼26 times higher than that of the traditional MIP with a nonporous structure.

Figure 8

(a) Heck coupling reaction. (b) Scheme for the synthesis of the PAF platform, MIPAF-0. (c) Preparation of MIPAF-1–4 with hydrolysis site. (d) Paration of MIPAF-5 and MIPAF-6 with product adsorption site. (e) Preparation of MIPAF-9 with substrate hydrolysis and product adsorption sites. (f) Hydrolysis experiments for various substrates in the presence of MIPAF-9. Stability of MIPAF-9 against temperature (g), pH (h), heavy metal ions (i), organic solvents (j), and cycles used (k). Reprinted with permission from ref (107). Copyright 2018, Wiley-VCH.

Using the hydrolysate (p-nitrophenol) as a template, a product adsorption complex is obtained by preassembly with zinc dimethacrylate. Similarly, MIPAF-5, MIPAF-6, MIPAF-7, and MIPAF-8 possess 10%, 20%, 30%, and 40%, respectively, of product adsorption complex. The BET surface areas of MIPAF-5, MIPAF-6, MIPAF-7, and MIPAF-8 are 507, 400, 283, and 191 m2 g–1, respectively. Regarding the capacity for p-nitrophenol, MIPAF-7 exhibits the largest uptake of 32.06 μg mg–1 among other MIPAF samples, which is a 4.6-fold increase relative to that of the traditional polymer-based MIP (5.7 μg mg–1). Both the substrate hydrolysis and product adsorption experiments make evident that PAF scaffolds will open up the interior spaces and enable the high utilization of imprinted sites.

As previously reported, the in situ separation of products from catalytic sites can effectively accelerate the catalytic process. In order to increase the catalytic rate, 30% substrate hydrolysis sites and 5% product adsorption sites were incorporated into a PAF scaffold to give MIPAF-9. The MIPAF-9 powder hydrolyzes the organophosphorus molecules and transports the hydrolysate (p-nitrophenol) from the catalytic site. The resulting catalytic system converts 17% of the paraoxon-ethyl in 2.5 h, which surpasses 14 times the rate of natural organophosphorus hydrolase (Flavobacterium sp. strain ATCC 27551). It is worth mentioning that, with the stable PAF as a scaffold, the MIPAF-9 sample exhibits excellent stability at high temperatures; in the presence of heavy metals, acids/bases, and organic solvents; and with cycling (Figure ). The integration of multiple functional units in the porous architecture facilitates the realization of efficient catalysis. In addition, the porosity and stability of the PAF platform provide more possibilities for practical applications.

Postmodification of Recognition Sites for Selective Halogenation

As previously observed, the topological structure of an imprinted complex is quite different from the segment of the porous framework. As imprinted sites are introduced into the architecture, the skeletal fragments are distorted, and the pore channels are blocked, which damages the integrity of the framework structure. Therefore, the specific surface area of PAF materials gradually decreases with increasing content of imprinted sites. After exceeding a certain upper limit (molar ratio of 40–50%) of the imprinted complex loading, the yields of the resulting solids decrease sharply. Maintaining the integrity of the porous architecture and increasing the amount of imprinted sites for high-performance molecular recognition is a goal that is currently being pursued.

Halide molecules are important ingredients in the fields of pharmaceuticals and chemical production. However, phenyl rings have multiple reactive sites, which introduces additional challenges to selectively synthesizing phenyl halides. To increase the effective payload of imprinted sites, the authors designed a fully fluorinated PAF solid, denoted as PAF-63.[108] PAF-63 exhibits unique advantages (Figure ), including the following: (i) The C–F bond has a higher energy than the C–H bond, helping to maintain the structural integrity and avoid the side reactions, and (ii) the hierarchical porosity facilitates the postsynthesis reactions for the recognition units.

Figure 9

Scheme for the preparation of fluorinated PAF solid, PAF-63 (a), and cyclodextrin modified PAF, CD-PAFs (b). Possible fragments of PAF-63 (c) and CD-PAFs (d). (e) Cavity diameters of α-CD, β-CD, and γ-CD. (f) Schematic diagram for the CD incorporated complex. Para sites are left in the CD incorporated complexes. (g) Several aromatic substrates (phenol, anisole, phenyl acetate, acetanilide, benzanilide, and 2-chloro-5-nitro-N-phenyl benzamide) with different kinetic diameters. Reprinted with permission from ref (108). Copyright 2017, American Chemical Society.

The PAF architecture was subsequently decorated with three CD molecules (α-CD, β-CD, and γ-CD) with internal diameters varying from 0.57 to 0.95 nm. The CD-modified PAF frameworks have hydrophobic cavities that can incorporate aromatic substrates due to the CD internal chamber. Under the protection of the hydrophobic cavities, the meta and ortho positions are embedded in the CD-PAF materials, and only the para position is exposed for halogenation reactions (Figure f). This study illustrates the postsynthesis modification method for obtaining a large number of imprinted sites while also maintaining the integrity of the PAF framework. This strategy prevents the collapse of the framework and the blockage of the mass transfer channels, greatly improving the molecular recognition performance of these materials.

Tackling an International Problem: Uranium Extraction

Currently, nuclear power provides 13% of the world’s electrical energy; and the International Atomic Energy Agency (IAEA) predicts that it will supply more than half of the electricity consumed globally in the next few decades. Uranium is an indispensable raw material in the nuclear energy industry. However, the uranium reserves that are easy to mine on land are about to be exhausted in the coming decades. Seawater contains the largest amount of uranium in the world and can help meet energy needs and ensure developments in the long run. Nevertheless, two major problems, i.e., the ultralow concentration (∼3.3 ppb) and presence of vast competing cations, hinder the efficient and effective extraction of uranium from the ocean. Recently, experts have developed several methods, such as exploiting electrodeposition methods, designing bioinspired nanotraps, utilizing bacterial adsorbents, etc., for addressing these challenges. Despite the great progress, their extraction performances must still be improved for practical utilization.

PAF materials featured by large specific surface areas and high stabilities show great potential in the enrichment of radioactive ingredients. Using PAF-1 (BET surface area ∼5600 m2 g–1) as the scaffold, Prof. Ma decorated the PAF-1 framework with amidoxime groups to obtain PAF-1-CH2AO, and Prof. Dai loaded polyacrylonitrile into its channels to prepare amidoxime-PAF-1.[109,110] Both materials reveal an excellent capability in terms of adsorption capacities and rates, but the selectivity still needs to be improved.

According to the specific coordination mode of uranyl ions, the authors synthesized a UO22+-imprinted complex consisting of salicylaldoxime, MAA, 4-VP, and UO2(NO3)2 via free assembly (Figure a). Through the Heck coupling reaction, the UO22+-imprinted complex together with 1,3,5-tris(4-bromophenyl)benzene and p-divinylbenzene was used to construct a series of PAF networks (Figure b,c). After removing template UO22+ ions, the carboxyl and pyridine groups in the PAF structure are fixed on the PAF structure through covalent bonds, whereas the salicylaldoxime units maintain their spatial positions through hydrogen bonding and π–π interactions.

Figure 10

(a) Structural units for MIPAF materials. (b) Synthetic routes and possible segments for traditional PAF. (c) Decorating UO22+-imprinted complex in MIPAF-11b–11d. (d) UO22+ ion capacity of MIPAF-11c against different interfering ions. (e) Comparison of selectivity and saturation uptake for UO22+ ion by various benchmark materials. Reprinted with permission from ref (111). Copyright 2018, Wiley-VCH.

The combination of MIT technology and porous materials effectively enables the highly efficient and selective adsorption of uranyl ions. MIPAF structures provide numerous accessible imprinted sites with a nearly 4-fold increase in the capacity relative to that of conventional MIPs. To our knowledge, the MIPAF-11c material has the highest reported selectivity (selective coefficient of greater than 746) of all uranium adsorbents (Figure ).[111] In addition, composite devices, including fibers, films, and coatings, can be fabricated by a flexible operation using the UO22+-imprinted PAF powder, enabling facile and convenient industrial utilization.

To improve the adsorption capacity while maintaining the high selectivity, the authors designed an imprinted complex (uranyl-specific bis-salicylaldoxime entity) by the ion coordination template strategy.[112] The bis-salicylaldoxime entity is tethered on the PAF skeleton through hydrogen bonding (Figure a). Regarding its spatial structure, the configuration of the uranyl-specific bis-salicylaldoxime entity is consistent with the fragment of PAF-1, and the imprinted sites maintain their structural integrity via π–π interactions (Figure b). Because the imprinted sites preserve the architecture and direction for uranium rebinding, the resulting adsorbent exhibits ultrahigh selectivity for uranium ions with a selectivity coefficient of greater than 113 (uranium/vanadium) and an outstanding capture capacity of 5.79 mg g–1 from real seawater in 56 days. Professor Daniel T. Sun wrote a critical article, stating that the porous aromatic framework material with uranyl-specific bis-salicylaldoxime entities has fascinating properties and is expected to solve the world problem of uranium extraction from seawater.[113]

Figure 11

Synthetic processes for double-carboxylic S-PAF-1. (a) Postsynthetic modification of the PAF-1: molecular coordination template strategy to produce UO22+ ion binding cavities (uranyl-specific bis-salicylaldoxime entity) with target shape complexity and binding affinity. Possible patterns of the uranyl-specific bis-salicylaldoxime entity fixed onto the PAF-1 according to the energy minimization optimization simulated by Materials Studio. (b) Comparison of coordination bond lengths for uranium and vanadium complexes. (c) Adsorption isotherm in different concentrations of UO22+ ion by MISS-PAF-1. (d) Comparison of ion selectivity and equilibrium uptake for various benchmark materials. (e) Selectivity against competing ions for MISS-PAF-1 in simulated seawater. Reprinted with permission from ref (112). Copyright 2019, American Chemical Society.

Having established a porous adsorbent with both high selectivity and capacity, conductive chains (phenylacetylene, PPA) were incorporated into the channels of porous adsorbents into PPA@MISS-PAF-1.[114] Under the asymmetrical alternating current electrochemical (AACE) method, the expended electrical field covers the micrometer-sized particles, which guide the migration and enrichment of uranyl ions (see Figure ). Based on the merits of both porous architecture and electrically driven motility, the superstructure shows faster kinetics 100–1000 times than the polymer-based adsorbents under physical diffusion. The resulting material realizes the ultrahigh performance for the uranium extraction, which captures 13.0 mg g–1 uranium in 56 days from natural seawater.

Figure 12

Chemical structure for phenylacetylene-doped MISS-PAF-1 and PPA@MISS-PAF-1, and physical processes occur under AACE extraction. Reprinted with permission from ref (114). Copyright 2020, Elsevier.

Summary and Perspective

Future work should focus on the following: (1) changing the hydrophilicity, hydrophobicity, and polarity of the pore surfaces of PAF materials to simulate the catalytic environments of biomimetic enzymes; (2) tuning pore structures to regulate the speed for mass transfer; (3) rationally assembling multiple imprinted sites to reasonably combine multiple functions; (4) utilizing conjugated building units with electrical or photophysical properties for chemo-/biosensing fields; and (5) imprinting macromolecules, such as biological enzymes, DNA, proteins, etc., for their monitoring and separation.

Safety Statement

No unexpected or unusually high safety hazards were encountered.