Metal–Organic Frameworks Meet Molecularly Imprinted Polymers: Insights and Prospects for Sensor Applications

Abstract

The use of porous materials as the core for synthesizing molecularly imprinted polymers (MIPs) adds significant value to the resulting sensing system. This review covers in detail the current progress and achievements regarding the synergistic combination of MIPs and porous materials, namely metal/covalent–organic frameworks (MOFs/COFs), including the application of such frameworks in the development of upgraded sensor platforms. The different processes involved in the synthesis of MOF/COF-MIPs are outlined, along with their intrinsic properties. Special attention is paid to debriefing the impact of the morphological changes that occur through the synergistic combination compared to those that occur due to the individual entities. Thereafter, the strategies used for building the sensors, as well as the transduction modes, are overviewed and discussed. This is followed by a full description of research advances for various types of MOF/COF-MIP-based (bio)sensors and their applications in the fields of environmental monitoring, food safety, and pharmaceutical analysis. Finally, the challenges/drawbacks, as well as the prospects of this research field, are discussed in detail.

1. Introduction

Molecularly imprinted polymers (MIPs), a special class of polymeric materials encompassing predesigned molecular recognition features, offer great potential in a wide range of applications including environmental monitoring, food safety, and clinical diagnosis.¹⁻⁴ Distinctly, MIPs are robust alternatives in contrast to the relatively fragile antibodies. Notably, the MIP is accomplished by the synthesis of a polymer in the presence of a given template molecule followed by the template removal step, imprinting the structural features, shape, size, and functionality, of the template/guest molecule within the resultant polymer.^5,6 Markedly, MIPs can selectively embed/capture the structural characteristics of small organic molecules and up to large molecules, such as proteins, viruses, and bacteria.^7,8 Certainly, their outstanding properties, including high chemical and mechanical stability and excellent recognition capability, prompted their exploration as biomimetic materials and the prospect of alternative biological receptors.^9,10 Prominently, MIPs offer numerous advantages, such as high adhesion of the polymer to the transducer surface, high selectivity toward the desired analyte, high physical robustness, ease of preparation, and low costs in contrast to biological receptors.¹¹⁻¹³ Poma et al. reported that the quality of synthesized MIPs is like that of monoclonal antibodies.¹⁴ Antibodies have disadvantages, such as a short lifetime (6–12 months), narrow pH range, reusability issues, high costs, and specific storage requirements. Moreover, MIPs are completely synthetic and animal-free alternatives, in contrast to the production of antibodies contingent upon the use of a host animal. In contrast, porous materials such as porous coordination polymers, also known as metal–organic frameworks (MOFs), are widely used for gas storage, membranes,^5,6 catalysis, drug release,^7,8 chemical sensors, and biosensors.^3,15 MOFs can be synthesized through the coordination of metal ions/clusters with organic linkers, which leads to periodic crystalline porous structures.¹⁶⁻¹⁸ The ability to design, fine-tune, and functionalize MOFs’ porous structures makes them suitable candidates for various applications.¹⁹⁻²¹ Covalent–organic frameworks (COFs) belong to the same spectrum of crystalline porous materials as MOFs. However, COFs comprise highly ordered networks built with organic linkers, connected through covalent bonds, and with light elements such as hydrogen, boron, carbon, nitrogen, and oxygen, unlike the metal clusters in MOFs.²² The unique features of COFs, such as their geometry and structures, allow them to predesignate either in two-dimensional (2D) or three-dimensional (3D) networks, which are composed of completely covalent bonds. The covalent bonds that form the frameworks make COFs thermodynamically stable.

Recently, researchers in this rapidly growing area have realized the various applications of multifunctional composites constructed from the controlled integration of MOFs/COFs with other materials.^23,24 Processing facilitated MOF composites to exhibit superior characteristics in terms of mechanical and chemical stability, selectivity, porosity, large surface area, scalability, and processability when compared to other pristine porous materials.²⁵ Thanks to these individual features of both MIPs and MOFs, their resultant synergistic combination composites have gained tremendous attention in the past decade.^26,27 Both MIPs and MOFs share a common feature in having pores and cavities as part of their geometrical signature. The use of MOFs as the core for synthesizing MIPs has been found to add significant value to the resulting sensing system, mainly in terms of the increased number of pores, improved sensitivity, and enhanced recognition sites. Riskin et al.²⁸ used bisaniline-cross-linked gold nanoparticles (AuNPs) followed by the electrochemical synthesis of a MIP based on gold p-aminothiophenol (PATP) as a functional monomer to detect 1,3,5-trinitrotoluene (TNT). The same strategy has subsequently been applied for detections of estradiol²⁹ and glyphosate.²⁹ Later, this strategy was applied using MOF/MIP to detect tetracycline³⁰ and aflatoxin B1.³¹ In another strategy, MOF has been successfully used as a core in MIP core–shell nanoparticles to be integrated into mass sensors.³²⁻³⁴ Furthermore, many successful applications for this combination strategy have been identified, especially in the solid-phase extraction process before the sensing step by different transducers.^15,35,36

The incorporation of MIPs into MOFs/COFs simultaneously combines the outstanding features of both excellent materials that provide selectivity and better adsorption capacity. Indeed, one of the main drawbacks of the MIPs building blocks is that their flexible skeletons in general carbon–carbon single bonds tend to deform and agglomerate into dense clusters, resulting in low selectivity and affinity, low adsorption capacity, and low mechanical stability.³⁷ To overcome these limitations, MOFs/COFs’ rigid skeleton can effectively allow the prevention of the distortion of the MIPs’ imprinted cavities. As a result, this combination of both materials (MOF/COF-MIP) allows the enhancement of the mechanical strength of the recognition sites as well as their selectivity. Furthermore, the combination of these materials can significantly enhance the accessibility to the imprinting sites thanks to their abundant pores with a larger specific surface area.³⁸

To prepare this review, we referred to general data published between 2000 and 2021 provided by the SCOPUS and the Web of Science databases. Notable growth in the number of documents published in the field of MOF/COF-MIP combinations, particularly in the last five years, was identified. This increase reflects the potential of and broad interest in this research field. To the best of our knowledge, only one mini-review has been published to date, namely that of Guo et al., which focused on MOF/COF-MIP-based chemical sensors; however, it only covered certain sensing aspects of MOF/COF-MIPs.²⁷ Several other reviews have partially covered MOF/COF-MIP combinations.^26,39,40 Thus, a need exists for a comprehensive review paper that critically discusses all the new aspects in this emerging research field. Therefore, we focused on the progress in MOF/COF-MIP combinations in the field of sensors in terms of the different synthesis methods used for these combinations and the application of the resulting materials in environmental monitoring, food safety, clinical diagnosis, and pharmaceutical analysis.

We describe the recent achievements of MOF/COF-MIP combinations by highlighting the following points:

• Strategies for the fabrication of MOF/COF-MIPs using chemical and electrochemical synthesis

• The properties of MOF/COF-MIPs

• The applications of MOF/COF-MIPs regarding environmental monitoring, food safety, and pharmaceutical analysis

• The current challenges and drawbacks associated with and the prospects of the use of MOF/COF-MIPs in sensing

2. Brief Description of MIPs, MOFs, and COFs

2.1. MIPs

Molecularly imprinted polymers are biomimetic synthetic materials that resemble antibodies in their structures since they have specific recognition sites called imprints that are capable of selectively binding the desired molecule with which they were templated during synthesis. From a very simple point of view, the functioning mechanism of these MIPs is like a “lock and key” process to selectively insert the molecule into its cavity. Contrary to real antibodies, MIPs are intrinsically stable in different media, low cost, and have a longer storage life.^6,41

In 1931, the subject of imprinting technology was opened by M. V. Polyakov when he discovered the creation of imprints left by a template molecule in porous silica particles for chromatographic applications.⁴² In 1949, 18 years later, F. H. Dickey polymerized silica gels in the presence of helianthin, then removed them from the silica gel, and, subsequently, re-exposed the gels to different species.⁴³ He observed a high selectivity for the species in which they were present during the polymerization compared to a silica gel reference which was polymerized without the presence of helianthin.

In the 1970s, molecular imprinting research turned away from silica toward organic polymers. G. Wulff⁴⁴ developed organic polymers in which the template was covalently bound to the polymer matrix. A significant revolution in the evolution of molecularly imprinted polymers was the introduction of the noncovalent approach by K. Mosbach’s team, at the beginning of the 1980s.^45,46 This method is based on the polymerization of monomers, whose functional groups interact with the template molecule through noncovalent interactions.

The key step in the molecular imprinting process is the formation of a complex before the polymerization step between the template molecule and the functional monomers. There are three distinct strategies, differing in the interactions, that lead to the formation of this complex including noncovalent, covalent, and semicovalent approaches.⁴⁷

Regarding the covalent approach, the template molecule and functional monomer complex is created by reversible covalent bonds. Therefore, before any polymerization, an advanced chemical synthesis that covalently grafts the template molecule on the selected monomer is required. After the polymerization, the template molecule is extracted by chemical cleavage, revealing cavities exposing functional groups which allow the recognition of the target molecule via covalent interactions. However, this approach is less simple to carry out and requires organic chemistry skills for the creation of the template-monomer complex. Moreover, the number of molecules that can be imprinted by this technique is smaller, since the chemical modification yield of templates is not always high. On the other hand, because of the longer time required by the covalent bonds to establish themselves, the rebinding kinetics is also lower overall than for the noncovalent approach.

In the noncovalent approach, the nature of interactions between the functional monomers and the template molecule is of hydrogen, weak electrostatic, ionic, van der Waals, or hydrophobic types. After the preparation of the MIP, the same type of weak interactions allows the recognition of the target molecule in the imprints created. Because of its simplicity of implementation, the wide choice of commercially available functional monomers, and the easy extraction of the template from the polymer matrix, the noncovalent approach is the most widespread one. The semicovalent approach combines the two previous approaches. The creation of the complex is achieved by covalent interactions, and molecular recognition with the target molecule is achieved by noncovalent interactions taking advantage of both methods. Nevertheless, this technique remains little used in the MIP community.

Different preparation strategies have been reported such as soft lithography or surface stamping,⁴⁸ phase inversion,⁴⁹ and synthesis. The first one involves the production of a template stamp on a polymer film, usually through self-assembly, and the stamp is then pressed onto a polymerizing film and kept in place until the film is solidified. Finally, the stamp is removed from the polymerized film by washing out it, and the binding cavities are left behind. The process is theoretically easy but requires noteworthy attention during the preparation. The second one consists of mixing a host polymer and the template in a compatible solvent, and the MIP is prepared by precipitation of the template-polymer complex after the addition of an incompatible solvent. The synthesis is the most used for the preparation of MIP and involves the polymerization of a monomer into a polymer using hydrolysis-condensation or radical polymerization methods. The synthesis of MIPs could be assisted by different techniques such as thermal heating,^12,50 photopolymerization,^51,52 microwave irradiation,⁵²⁻⁵⁴ sonochemistry,^3,55 and electrochemistry,^56,57 among others. There is also another way to talk about synthesis strategies of MIP including bulk polymerization,^58,59 precipitation,^60,61 emulsion polymerization,^21,40 suspension polymerization,^51,62 and reversible addition fragmentation chain transfer polymerization.^63,64 The choice of the preparation strategy depends on several factors such as the template molecule, final properties of MIP, the application of MIP, and availability of instruments and reagents.

2.2. MOFs and COFs

MOFs are a class of porous crystalline materials with organic and inorganic constituents, whereby inorganic metal ions/clusters are linked by organic ligands to form a rigid periodic network (Figure 1a). MOFs possess excellent potential for a plethora of applications for addressing challenges associated with energy and environmental sustainability.⁶⁵ MOFs have thus attracted a great deal of attention due to their wide scope of application, with researchers focusing on the synthesis and chemistry of these frameworks to develop an extensive variety of aesthetically interesting materials ranging from 1D to 3D materials with tunable metrics, porosity, and organic functionality.⁶⁶ MOFs’ unique structures that are composed of both organic and inorganic components in a rigid periodic networked structure are not readily accessible in other conventional porous materials, e.g., purely inorganic zeolites. Various combinations of metal ions/clusters and ligands can result in the construction of a nearly infinite range of MOF structures. Hence, a judicious and logical approach to the synthesis of predetermined MOF structures that includes the identification of target building blocks and the acquisition of knowledge of targeted networks must be adopted. Such a logical approach to synthesizing predetermined/designed MOF structures and crystalline solid-state frameworks is often called reticular synthesis.⁶⁷ It was shown that the design of an extended network can be realized by starting with well-defined and rigid molecular building blocks that will maintain their structural integrity throughout the construction process, unlike the retrosynthesis of organic compounds.⁶⁸

Figure 1.

(a) Representative schematic of the modular synthesis of MOFs. Reprinted with permission from ref (70). Copyright 2019, Wiley. (b) Schematic highlighting the importance of synthesis to MOF tenability.

The final structures and properties of MOFs are highly influenced by the synthetic methods used as well as conditions such as pressure, temperature, reaction time, solvents, and pH.⁶⁹ The reaction temperature is a key parameter that determines the kind of reaction setups to be employed.

Some common MOFs, such as MOF-5, HKUST-1, and ZIF-8, can even be obtained at room temperature by direct precipitation through the direct mixing of starting materials.⁶⁶ However, higher reaction temperatures (of up to 250 °C) are required to achieve suitable reaction rates and crystallinity for most systems. The control of temperature can be achieved through various energy sources, such as electrical heating, electromagnetic radiation, electric potential, and mechanical waves/ultrasound. Hence, several synthetic methods and approaches can be adopted to synthesize MOFs, such as hydrothermal/solvothermal, microwave-assisted, electrochemical, slow diffusion, mechanochemical, ultrasonic, and heating.⁶⁹ The particle size, distributions, and morphologies, in turn, determine the properties of materials, depending on the chosen synthesis route (Figure 1b). The sonochemical²³ and microwave-assisted⁷¹ synthetic methods are recommended to obtain nanocrystalline MOFs for biomedical applications. From another perspective, the possibility of directly growing MOF structures on a sensing substrate as thin films using the solvothermal synthesis technique has been achieved. For instance, MOFs such as MFM-300,⁷² NDC-Y-fcu-MOF,⁷³ and fum-fcu-MOF⁶⁵ were directly grown on interdigitated electrodes through the solvothermal method.

The stability of MOFs is paramount for real-time applications in the biomedical application field, especially drug delivery.²⁰ The stability of MOFs can be predicted from the metal–ligand bond strength, which forms the framework of MOFs. The hydrophobicity of the pore surface, organic ligands, operating environment, and metal–ligand coordination geometry are various factors that affect the stability of MOFs.⁷⁰ As per the hard/soft acid/base principle, MOFs are synthesized using a carboxylate ligand, which is a hard Lewis acid with heavy-valent metal ions (Ti, Zr, Al, Fe, and Cr), which can form stable MOFs.⁷⁴ In addition, soft azolate ligands such as imidazolates, pyrazolates, triazolates, and tetrazolates, which are soft Lewis base ligands with soft divalent metal ions (Zn, Cu, Ni, Mn, and Ag), can also form stable MOF structures. MOFs with these combinations can provide inherent stability. In addition, a postsynthetic modification can be adopted in MOF structures to improve their hydrophobicity and stability.⁷⁰ Through synthesis, various MOF attributes are tunable based on the required application, including physical and chemical properties, charge conduction, scalability, processability, porosity, and surface area.⁵¹

In the case of COFs, the covalent bonds that form the frameworks make them thermodynamically stable. Like MOFs, the functionality and porosity of COFs are synthetically tunable, which facilitates their wide application.^75,76 The design principles of COFs are different from those of cross-linked polymers and linear polymers. Highly ordered predesignable structures are obtained through topology diagram-directed polymer growth and geometry matching among the monomers (Figure 2). The topology design diagram describes the direction of covalent bond formation, which can be a design manual for polymer backbone growth to predesign the COF structures.⁷⁵ The design principles and methodologies for synthesizing building blocks for 2D (Figure 2a) and 3D COFs (Figure 2b) have been extensively discussed in a recent review.⁷⁵ Once the needed building blocks have been identified, the proper choice of the reaction media and conditions, which are crucial for synthesizing highly ordered and crystalline COFs, must be considered. Similar to MOF synthesis procedures, various synthetic routes have been explored for COF synthesis, including solvothermal, microwave-assisted, sonochemical, ionothermal, mechanochemical, and light-assisted methods.⁷⁷ As outlined in Figure 2c, each of these techniques has pros and cons. The solvothermal method is time-consuming due to the long reaction time, but it is a commonly used technique for COF synthesis. Sonochemical synthesis is fast and cost-effective and can be considered a green method. Microwave-assisted synthesis is a faster route and has significant potential to be scaled up. Ionothermal routes are chosen when high temperatures are required. In addition, green synthesis routes enable the environmentally friendly production of COFs.⁷⁸

Figure 2.

Topological diagrams for designing (a) 2D and (b) 3D COFs. Reprinted with permission from ref (75). Copyright 2014, ACS. (c) Properties and highlights of different synthesis methods of COFs.

3. Strategies for the synthesis of MOF/COF-MIPs

This section discusses all the techniques reported to date on the synthesis of MOF/COF-MIP combinations. Most of the aspects related to the synthesis are explained in detail and are presented in Table 1. Here, we focus not only on chemical synthesis but also on electrochemical synthesis.

Table 1. Features of the Chemical Synthesis of MIP/MOF Combinations^a.

Supporting material	Mechanism polymerization	Method	Analyte	Solvent/monomer/cross-linker/initiator or catalyst	Polymerization conditions	Washing solution
UiO-66 MOF58	FRP	Precipitation	Tetracyclines	Water/acrylic acid and methacrylic acid/N,N-methylenebisacrylamide (BIS)/APS	Room temperature, 20 h	Methanol/acetic acid (9:1, v/v)
UiO-66-NH290	FRP	Precipitation	OXA	Acetonitrile-methanol (2:1 V/V)/MAA/EGDMA/AIBN	70 °C, 24 h	0.20 M NaOH
NH2-MIL-101(Cr)79	FRP	Precipitation	SXD	Acetonitrile-ethanol (1:2, v/v) /MAA/EGDMA/AIBN	60 °C, 24 h	Methanol/acetic acid (9:1, v/v
MIL-1080	FRP	Precipitation	Coumarin-3-carboxylic acid	Ethanol/MAA/EDMA/AIBN	88 °C, 6 h	–
Eu(BTC)91	FRP	Precipitation	Lincomycin	Methanol and DMF/4-VP/EGDMA/AIBN	60 °C, 2 h	Water
UiO-66/phosphomolybdic acid hydrate80	FRP	Precipitation	H2O as a dummy template for H2S	Acetonitrile and ethyl acetate/acrylamide/EGDMA/benzoyl peroxide and N,N-dimethyl aniline	4 °C, 24 h	Methanol
ZIF-892	FRP	Precipitation	Furosemide	Methanol-acetonitrile (V/V 1:2)/MAA/EGDMA/AIBN	65 °C	0.25 M NaOH
MIL-101 @SiO2@MPs75	FRP	Precipitation	JEV	DMF/zinc acrylate/EGDMA/AIBN	5 h	Methanol and acetic acid (9:1, v/v)
GO/ZIF-8 MOF34	FRP	Precipitation	Progesterone	Acetonitrile and DMF (v/v, 3/1)/MAA/EGDMA/AIBN	60 °C, 12 h	Methanol:acetic acid:water (4:1:1, v/v)
UiO-66-NH2@GMA 5 (glycidyl methacrylate)84	FRP	Free radical poly precipitation	Quercetin	Ethanol/acrylamide/EGDMA/AIBN	7 h	Methanol/acetic acid (4:1, v/v) at 80 °C in a water bath
UMCM-1 MOF33	FRP	Precipitation using deep eutectic solvent	Di-isobutyl phthalate	DES/MAA/EGDME/AIBN	60 °C, 12 h	Acetonitrile:methanol (4:1)
Magnetic COFs conuclear structure with CDs58	Hydrolysis coupled with condensation	Reverse microemulsion	TNP	n-Hexanol, cyclohexane, and water/APTES/TEOS/ammonia	RT, 10 h	Methanol/acetic acid (9:1, v/v)
Fe3O4-NH221	Condensation reaction	–	Anthocyanins dummy template for C3G	Ethanol/DAAQ/trigonal building blocks (Tp)	RT, 24 h	MeOH/acetic acid (9:1, v/v)
UiO-6682	Hydrolysis and then condensation	Sol–gel	Tyramine	Ethanol/APTES/TEOS/ammonia	RT, 34 h	Ethanol and water (1:1)
HKUST-1@TEOS and APTES83	Hydrolysis and then condensation	Sol–gel	TBBPA	Ethanol/APTE/TEOS/acetic acid	60 °C, 24 h	Methanol/acetic acid (9/1, v/v)
MOFs35	Hydrolysis and then condensation	Sol–gel	Cyhalothrin	THF/APTES/TEOS/acetic acid	60 °C, 20 h	10% acetic acid in methanol
Ni@MIL-100(Fe)93	Hydrolysis and then condensation	Sol–gel	Hydroxychloroquine	Ethanol/APTES/TEOS/acetic acid	50 °C, 6 h first; 60 °C, 24 h; and finally 30 °C, 18 h	Methanol:acetic acid solution (9:1, v/v)
MOF-17757	–	Sol–gel	S-Amlodipine	Acetonitile/MAA/TEOS/acetic acid	60 °C, 20 h	Acetonitrile
NH2-MIL-53(Al)61	Hydrolysis and then condensation	Sol–gel	Cyclodo-decanyl-2,4-dihydroxybenzoate (CDHB), an analog of ZON	Ethanol/APTES/TEOS/ammonia	RT, 24 h. The MMIP was calcined in a vacuum oven at 130 °C, 2 h	Ethanol and ammonia (9:1, v/v)
MIL-10194	FRP	–	Hepatitis A virus	THF/APTES/N,N′-methylenebisacrylamide (MBA)/APS	60 °C, 24 h	Methanol/acetic acid (9:1, v/v)
Zr-MOF@graphene51	FRP	UV-induced polymerization directly on the surface of the electrode	Ketamine	Methanol and acetonitrile/MAA/EGDMA/AIBN	Exposure of UV wavelength at 380 nm, 3 h	1:9 (v/v) of acetic acid/methanol
ZIF-887	FRP	Pickering emulsion polymerization	2,4-Dichlorophenoxyacetic acid	Toluene/4-VP/EGDMA/AIBN Dodecane as the phase transfer agent was added afterward to form the oil phase	65 °C, 16 h, N2 for 10 min	Methanol/acetic acid (7:3, v/v)
Carbon dots-embedded COFs26	FRP	One-pot surface-imprinting synthesis	Tryptamine	–/MAA and AM/EGDMA/AIBN	60 °C in a water bath, 20 h	Methanol
QDs (CdSe/ZnS)-grafted COFs (TpPA)26	Hydrolysis and then condensation	One-pot reverse microemulsion strategy	Quinoxaline-2-carboxylic acid	Ethanol/APTES/TEOS/ammonia	24 h	Methanol
QDs (CdSe/ZnS)-grafted COFs (TpPA)26	Hydrolysis and then condensation	One-pot reverse microemulsion polymerization bulk polymers	Tyramine	Ethanol/APTES/TEOS/ammonia	20 h	Methanol
Carbon nanodot (CN)-grafted COFs (TPPA COFs)51	Hydrolysis and then condensation	One-pot room-temperature synthesis, reverse microemulsion polymerization	4-ethylguaiacol (4-EG)	–/APTES/TEOS/ammonia	Overnight, 48 h	Methanol
QD-grafted COFs CdSe/ZnS TpPa15	Hydrolysis and then condensation	Reverse microemulsion, one-pot surface-imprinting	Ferulic acid	Ethanol/APTES/TEOS/ammonia. Triton X-100 and cyclohexane were used as surfactant and oil phase, respectively.	Overnight	Acetone
UiO-66-NH240	FRP	–	Ofloxacin	Methanol/acetonitrile (1/1, v/v)/MAA/TRIM/AIBN	60 °C, 24 h	Methanol/acetic acid (9/1, v/v)
CdSe/ZnS QD-grafted TpPa COFs23	Hydrolysis and then condensation	Reverse microemulsion	Bovine hemoglobin	–/TEOS/APTES/ammonia. Cyclohexane served as the continuous phase, and Triton X-100 was the surfactant.	Overnight	PBS
HKUST-126	–	Precipitation-liquid crystalline	Capecitabine	Toluene and acetonitrile/4-Methyl phenyl dicyclohexyl ethylene/EGDMA/AIBN	60 °C, 24 h	Methanol

^aAbbreviations: FRP: Free radical polymerization; APS: ammonium persulfate; OXA: oxaliplatin; MAA: methacrylic acid; EGDMA: ethylene glycol dimethacrylate; AIBN: azobisisobutyronitrile; SXD: Sanhuang Xiexin decoction; 4-VP: 4-vinylpyridine; DMF: dimethyl formamide; Eu(BTC)-MIP: europium metal–organic framework coated with molecularly imprinted polymers; JEV: Japanese encephalitis virus; GO/ZIF-8 MOF: graphene oxide, zeolite imidazolate framework 8; UiO-66-NH₂: amino-functionalized zirconium-based metal–organic framework; APTES: (3-aminopropyl) triethoxysilane; TEOS: tetraethoxysilane; CDs: carbon dots; TNP: 2,4,6-trinitrophenol; Fe₃O₄-NH₂: amino-functionalized magnetic nanoparticles; DAAQ: 1,5-diaminoanthraquinone; MeOH: methanol; TBBPA: tetrabromobisphenol A; THF: tetrahydrofuran.

3.1. Chemical Synthesis

The synthesis of MOF/COF-MIP composites begins with the proper choice of MOFs or COFs, which serve as one of the primary constituents. Several comprehensive reviews have discussed the design and synthesis of MOFs^{51,66,67,69,70} and COFs.^75,76 Nevertheless, here, we briefly highlight the reported techniques and factors to be considered for MOF and COF synthesis; these also apply to the synthesis of MOF/COF-MIP hybrids.

3.1.1. Precipitation Method

The precipitation polymerization method is considered the simplest strategy and is a widely used method for the rapid synthesis of high-yield MOF/COF-MIPs. Its advantages are the elimination of the need for stabilizers, simplicity of preparation, compatibility with cross-linking agents, and the elimination of grinding and sieving steps, all of which facilitate large-scale production. In this method, synthesis is conducted in the presence of a large amount of an appropriate solvent. During the synthesis process, the polymer chains grow and become insoluble in the medium. As a result, spherical particles at the nano- or microscale are obtained depending on the polymerization parameters.

The adsorption and selectivity performance of MOF/COF-MIP composites highly depend on several synthesis factors, including the type of solvent, monomer, template:monomer ratio, and cross-linker. Pan et al.⁷⁹ published a study of these factors for baicalein, berberine, and emodin using the NH₂-MIL-101(Cr) MOF. Furthermore, Pan et al.⁸⁰ reported the preparation of a MOF/MIP composite using a structural analog of the targeted analyte as a dummy template using precipitation polymerization. Indeed, the strategy based on a dummy template can overcome several problems, such as the high cost of the target, risk of template leakage, and possibly harmful operation.

Pan et al. developed a MOF-MIP composite with a core–shell structure for H₂S detection. H₂O was used as an analog template to avoid the toxicity and low stability of H₂S. The UiO-66 MOF modified by phosphomolybdic acid hydrate was used as a core material. Furthermore, the authors used a mixture of acetonitrile and ethyl acetate as a porogen solvent, ethylene glycol dimethacylate (EGDMA) as a cross-linker, and benzoyl peroxide/N,N-dimethyl aniline as initiators. Toward the end of the synthesis process, the optimal polymerization conditions were a volume ratio of acetonitrile to ethyl acetate of 1:1, a molar ratio of H₂O/acrylamide/EGDMA of 1:4:10, and a polymerization time of 24 h.⁸⁰

3.1.2. Sol–Gel Method

The sol–gel synthesis of MOF/COF-MIPs involves the use of organo-silanes as metal alkoxide molecular precursors, which are formed by hydrolysis and condensation in the presence of a template molecule and highly cross-linked materials.⁸¹ The monomers and cross-linkers used in this method differ from the ones used in the traditional method based on radical polymerization, which involves the use of vinyl monomers and cross-linkers. The available monomers for preparing MOF/COF-MIP composites using the sol–gel method are aminopropyltriethoxysilane (APTES), N-(aminoethyl)-3-aminopropyl-triethoxysilane (AATES), 3-isocyanato-propyltriethoxysilane (ICPTES), and methyltriethoxysilane (MTES). Several cross-linkers have been used to prepare MIPs with the sol–gel method, such as tetraethoxysilane (TEOS), tetramethoxysilane (TMOS), 1,2-bis(triethoxysilyl)ethene (BTEE), 1,4-bis(triethoxysilyl)benzene (BTEB), bis(triethoxysilyl)methane (BTEM), and 1,2-bis(triethoxysilyl)ethane (BTEA), but TEOS is the most frequently used. The reaction catalysis could be performed in acidic or basic mediums. Yao et al.⁸³ have reported the preparation of an ultrastable MOF/MIP composite through a sol–gel method for the determination of tyramine. The template and functional monomer APTES were dissolved in ethanol and stirred for 30 min. Then, the cross-linker TEOS was added, followed by aqua ammonia, which catalyzed the formation of a highly cross-linked material. The synthesis process took 24 h for polymerization and was aged for 10 h. The removal of the template was performed using water:ethanol (1:1). The sol–gel method was also adopted to prepare MIP composed of a copper-based MOF (HKUST-1) on paper support to selectively uptake tetrabromobisphenol-A (TBBPA).⁸³ The procedure reported was performed in three steps. The first step was the activation of filter papers with 3-(trimethoxysilyl)propyl methacrylate (γ-MAPS) to improve the hydrophobicity of the material. The second step consisted of modification with the MOF (HKUST-1) combined with APTES and TEOS. Finally, the third step was the growth of MIP on the surface paper-based MOF with a sol–gel method using APTES as a monomer, TEOS as a cross-linker, and acetic acid to catalyze the reaction. The prepolymerization solution was incubated with the modified papers at 60 °C for 24 h. The resulting MOF-MIP-based paper was washed with ethanol and methanol/acetic acid (9/1, v/v) to remove the unreacted constituents and templates and then dried in a vacuum oven at 60 °C. Liang et al.⁸⁵ have reported a straightforward procedure for the preparation of MOF-MIP by mixing cyhalothrin (template) with the MOFs and APTES in tetrahydrofuran (THF) for 60 min at room temperature. Then TEOS and acetic acid were added and maintained at 60 °C for 20 h. A Soxhlet extractor was used with methanol and acetic acid (9:1) to remove the template. Finally, the resulting material was dried overnight at 80 °C. Mathieu-Scheers et al. have reported the preparation of a core–shell hybrid of MOF/MIP (Figure 3B). S-amlodipine (S-AML), MAA, TEOS, MOF-177, and acetic acid were used as a template, monomer, cross-linker, core support, and catalyst, respectively. The MOF-177 was composed of triangular 1,3,5-benzenetribenzoate units and octahedral Zn₄O(−COO)₆ to construct a 3D extended network. The authors studied several synthesis parameters, including the amount of template molecule, cross-linker, and catalyst, as well as the polymerization and sonication times, for the prepolymerization solution to obtain an adsorption capacity of 1.31 mmol·g^–1 and an imprinting factor of 5.79. Simply, the templates (S-AML, MAA, and MOF-177) were introduced in acetonitrile and stirred for 1 h. Thereafter, TEOS and acetic acid were added and stirred for 30 min. After 10 min under N₂ purging, polymerization was performed at 60 °C for 20 h. The last step consisted of removing the template using acetonitrile.⁵⁷

Figure 3.

(A) Schematic representation of MIPs based on ZIF-8 stabilized Pickering emulsion polymerization. Reprinted with permission from ref (87). Copyright 1999, Elsevier. (B) Fabrication of a core–shell sol–gel hybrid MIP based on MOFs. Reprinted with permission from ref (57). Copyright 2019, Elsevier. (C) Preparation of magnetic COF/MIP. Reprinted with permission from ref (21). Copyright 2020, Elsevier. (D) A novel ketamine electrochemical sensor was constructed by UV-induced polymerized KT-imprinted membranes on the surface of graphene and a MOF nanocomposite modified screen-printed electrode using methacrylic acid as the functional monomer. Reprinted with permission from ref (51). Copyright 2017, Elsevier.

3.1.3. MOF Deep Eutectic Solvent-MIPs

Mirzajani et al. have discussed the preparation of MOF deep eutectic solvents (DESs) and MIPs in two different forms, namely monolithic and hollow fibers using UMCM-1 as the MOF material, DES and DMF as the solvents, MAA as the monomer, EGDMA as the cross-linker, and AIBN as the and initiator.³⁴ DESs are prepared by forming a complex between a hydrogen-bond donor with functional groups and a quaternary ammonium salt such as choline chloride. The authors used DES as a substrate for MOFs to increase their stability, enhance the location of the target analyte, improve the binding capacity, and accelerate the kinetic adsorption.

3.1.4. One-Step Synthesis of COFs along with Imprinted Cavities

In contrast to other synthesis strategies, the condensation method allows the synthesis of COFs accompanied by imprinted sites in one step. The condensation reaction occurs at room temperature, which effectively ensures the stability of the heat-sensitive templates in the formation of the COF-MIP. Furthermore, the building units of COFs can be used as functional monomers of MIPs to extend the type of monomers that can be used to prepare MIPs. Moreover, condensation polymerization does not require the use of initiators that are toxic and require the bubbling of inert gas. Recently, Zhao et al.²¹ have reported a one-step synthesis strategy for a COF-MIP material on the surface of magnetic nanoparticles (MNPs) at room temperature (Figure 3C). First, MNPs were prepared in the presence of 1,6-hexanediamine to functionalize their surface with amino groups for covalent linking with COFs through the Schiff base reaction. Then, seven magnetic COF-MIPs were synthesized using different monomers (linear building blocks) containing amino groups, such as p-phenylenediamine (Pa-1), benzidine (BD), 4,4″-diamino-p-terphenyl (DT), 3,3-dihydroxybenzidine (DHBD), 2,6-diaminoanthraquinone (DAAQ), 4,4′-diaminodiphenyl ether (ODA), 1,5-diamino-4,8-dihydroxyanthraquinone (DADHAQ), and 1,3,5-triformylphoroglucinol (Tp), as cross-linkers. Briefly, a mixture of a template and a monomer in ethanol was stirred for 4 h for the prepolymerization. Subsequently, the amino-functionalized MNPs were added and sonicated for 20 min. Thereafter, the cross-linker (Tp) was added, and the mixture was maintained under stirring for 30 min. Then, scandium(III) triflate was added dropwise as a metal triflate⁸⁵ to the mixture, followed by stirring for 24 h at room temperature. The resultant material was magnetically isolated and washed to remove the unreacted components, and the template was removed with methanol and acetic acid (9:1, v/v). This strategy proved to be an efficient method for the synthesis of a COF-MIP composite.

3.1.5. Liquid Crystalline MOF-MIPs

Conventional methods require a large amount of cross-linking agent (80 to 90%) to limit the relaxation phenomenon of the polymer backbones and obtain an efficient molecular recognition. However, the greater amount of cross-linker required for MIP synthesis could present several drawbacks related to the imprinted network. The resulting polymer is highly rigid and thus reduces the interaction between the template and the imprint, thereby impeding the extraction mechanism. As a result, only a part of the imprinted sites remained available.⁸⁶ Liquid crystalline MIPs are a new type of robust MIP that use liquid crystalline as an auxiliary monomer. The strong coupling interaction between the mesogenic groups of liquid crystalline and the polymer allows the shape of the imprints to be maintained due to the substitution of partial chemical cross-linking with physical cross-linking in the synthesis process. In this context, Coudert et al. developed a new liquid crystalline MIP based on HKUST-1 to combine the high selectivity of MIPs and the porous structure of HKUST-1. Here, 4-methyl phenyl dicyclohexyl ethylene was used as a liquid crystalline monomer to increase the solvent-responsive floating of the composite.²⁶ Moreover, the synergistic combination demonstrated higher relative bioavailability properties, which could be useful for drug delivery applications. The potential for oral administration was confirmed by the properties of the controlled release of capecitabine. With the use of liquid crystal MIP on HKUST-1, the latter’s poor stability in water was significantly improved.

3.1.6. UV-Triggered Polymerization Method

In addition, the polymerization of a MIP layer onto the surface of an electrode already modified with a MOF material by drop-casting has been reported by Lahcen et al. as a strategy for preparing MOF-MIP composites.⁵¹ The author’s approach was based on the preparation of a polymerization solution (solvent, monomer, template, cross-linker, and photoinitiator), then a small volume is dropped onto the surface of the electrode modified with MOF, and the polymerization occurs under UV exposure (Figure 3D). The authors reported optimization of the functional monomer, including MAA, acrylamide (AM), 2-(allylmercapto)nicotinic acid (ANA), and 2-acetamidoacrylic acid (AAA), with UV spectrophotometry being used to investigate the interaction between the template (ketamine) and monomer. The interaction led to absorption peak shifts of the molecule template with a hyperchromicity effect after the addition of the monomers because of hydrogen bonding. MAA exhibited a strong shift, indicating that it interacted powerfully with the template molecule compared with the other monomers. It is well-known that MAA is an acidic monomer that contains an hydrogen-bonding donor and can form a stable complex with alkaline and hydrogen-bonding receptor molecules such as ketamine. The molar ratio of the template, monomer, and cross-linker affects the imprinting efficiency and several specific cavities in the MIP film. The optimal molar ratio of ketamine, MAA, and EGDMA is 1:4:40.

3.1.7. Emulsion Method

3.1.7.1. Reverse Microemulsion

Ji et al.⁷¹ developed a novel reverse microemulsion strategy coupled with surface imprinting technology for the synthesis of quantum dot (QD)-grafted COF and MIPs for protein optosensing. Briefly, triton X-100 (as a surfactant) and cyclohexane (as a continuous phase) were stirred for 15 min. Then, the QDs, TEOS (cross-linker), ammonia (catalyst), APTES (monomer), and 1,3,5-triformylphloroglucinol-P-phenylenediamine (TpPA) TpPA-COF were added. After 2 h, bovine hemoglobin (template) was added, and the mixture was stirred overnight for polymerization. The template was removed with phosphate-buffered saline (PBS). The TpPA-COF reacted with QDs through the Schiff base reaction. Here, the APTES played a crucial role in binding the protein onto the surface of the silica material by noncovalent interactions. COFs were used to modify the QDs and provided a good support platform with high hydrothermal and chemical stabilities as well as an ultrahigh-specific surface area. Similar strategies based on reverse microemulsion were successfully applied later for other templates.^15,51,88

3.1.7.2. Pickering Emulsion Polymerization Method

This process involves the use of solid particles that serve primarily as Pickering stabilizers instead of traditional surfactants.⁸⁹ This method has been widely used for the preparation of high-yield and monodisperse particles that have uniform sites with effective binding capacity on the surface. It can resolve the problem of the deep embedding of templates using other traditional methods.²⁴ Recently, Marty et al. have described the synthesis of MIP-zeolite imidazolate framework-8 (ZIF-8) using Pickering emulsion polymerization. The ZIF-8 material played two roles, namely as a Pickering emulsion stabilizer and as a sacrificial material of MIPs. The template 2,4-dichlorophenoxyacetic acid was dissolved with the functional monomer 4-vynilpyridine in toluene and oscillated for 3 h. Then, the cross-linker EGDMA, initiator AIBN, and phase transfer agent dodecane (oil phase) were added. The water phase was obtained by dispersing the ZIF-8 material in water. Thereafter, the oil phase was poured into the water phase under high-speed stirring for 2 min to form a ZIF-8 stabilized Pickering emulsion. After bubbling the emulsion with N₂ for 10 min, polymerization occurred at 65 °C for 16 h without stirring. A mixture of 1.0 M HCl and methanol (4/1, v/v) was used to remove the self-sacrificial material (ZIF-8 nanoparticles) by stirring for 2 days. Finally, the template was removed with methanol/acetic acid (7:3, v/v).⁸⁷

3.2. Electrochemical Synthesis of MOF/COF-MIP Combinations

Electrochemical synthesis is an attractive strategy for the rapid synthesis of MIPs.^95,96 In addition, the electrosynthesis technique is exceptionally well suited to MIP preparation, especially for certain analytes, such as proteins,⁹⁷ bacterial cells,^55,98 and viruses.⁹⁹ Most electropolymerizable monomers can be electrodeposited from aqueous solutions where target analytes keep their natural conformation. The electrosynthesis of MIPs usually does not require an external initiator for the polymerization reaction.¹⁰⁰ Precise control over electrosynthesis enables the fine-tuning of the imprinted polymer layer’s thickness, which is particularly critical for surface imprinting with targeted analytes.

A MOF-imprinted polymer is deposited on functionalized nanoparticles by electropolymerization in the presence of the template. AuNPs are one of the most used metal nanoparticles for enhancing conductivity, as they can easily bind with thiol and amines. Duan et al. modified the surface of GCE with AuNPs supported on a Ti-benzeneaminodicarboxylate MOF (Au/NH₂-MIL-125(Ti)) for the detection of bovine serum albumin (BSA). They used BSA as a template for the electropolymerization of l-cysteine to prepare a MIP.¹⁰²

In another study, Rawool and Srivastava used Cu-MOF and mesoporous carbon to modify the surface of a glassy carbon electrode (GCE) for the detection of rifampicin (RIF) and isoniazid (INZ) using a dual-template MIP. Pyrrole was used as a monomer and RIF and INZ as templates to fabricate a dual-cavity MIP for the simultaneous detection of both analytes using adsorptive stripping differential pulse voltammetry (AdDPV).¹⁰³ In the same context, Iskierko et al. coated the surface of an Au electrode through the electropolymerization of 2,2′-bithiophene-5-carboxylic acid. Then, MOF-5 was grown on the carboxyl-functionalized Au-surface, and p-bis(2,2′-bithien-5-yl)methyl-o-catechol was used as a monomer due to its good interaction with NGAL protein (Figure 4B).²⁰⁰ More studies have reported electropolymerized MIPs on MOF-based sensors for aflatoxin B1,³¹ dopamine,⁶¹ tetracycline,³⁰ and l-phenylalanine.⁵⁰ Manousi et al. used electropolymerized p-aminobenzoic acid to prepare MIPs on molybdenum disulfide (MoS₂)/amino-functionalized carbon nanotubes (NH₂-CNTs)@COFs for sulfamerazine detection.⁴⁰ In electrosynthesis, the thickness of the imprinted films can be easily controlled by selecting the appropriate electrochemical parameters, including the scan rate, potential range, applied potential, and the number of cycles. So far, the different electrochemical strategies reported are summarized in Table 2.

Figure 4.

(A) Schematic illustration of (i) the fabrication process of CTpBD-Au hybrid, (ii) the preparation and detection process of developed MIECL, (iii) the light-emitting mechanism of an ECL system, and (iv) the quenching mechanism of ECL. Reprinted with permission from ref (61). Copyright 2011, Elsevier. (B) Electrochemical synthesis of a MIP/MOF combination at gold electrode. Reprinted with permission from ref (103). Copyright 2019, Elsevier.

Table 2. Summary of the Features of MOF/COF-MIP Sensors Prepared Using Electrochemical Methods^a.

Electrode configuration	Functional monomer	Analyte	Linear range	Detection method	LOD	Real sample
MIP/Au@Cu-MOF/N-GQDs/GCE90	Aniline	Patulin	0.001–70.0 ng mL–1	DPV	0.0007 ng mL–1, 0.7 pg mL–1	Apple juice
MIP/pTH/Au@ZIF-67/GCE21	Aniline	Tyr	0.01–4.0 μM	DPV	0.79 nM	Human serum
Au-NPS/MIP-MMOF/TC30	Aniline	TC	224 fM–22.4 nM	DPV	0.22 fM	Honey samples
MIP/Fe3O4@ZIF-67/Au71	Pyrrole	GA	6.0–600 pM	DPV	0.297 pM	Drinks
MIP/GO@COF/GCE40	Pyrrole	SDZ, AP	0.5–200 μM (SDZ), 0.05–20 μM (AP)	DPV	0.16 μM (SDZ), 0.032 μM (AP)	Pork, chicken
MIP-PB-MIL-CNTs/GCE61	Pyrrole	E2	10 fM–1.0 nM	DPV	6.19 fM	Water samples
Cu-MOF/MC/MIP/GCE102	Pyrrole	RIF, INZ	0.08–85 μM	CV	0.28 nM (RIF), 0.37 nM (INZ)	Drugs, human serum, and urine
GCE/Fe(III)-MIL-88B-NH2@ZnSeQDs/Ab200	Dopamine	SCCA	0.0001–100 ng mL–1	ECL	31 fg mL–1	Human serum
Zr-MOF/MIP105	pABA	DFC	6.5–1500 μM	DPV	0.1 μM	Pharmaceutical samples
ZIF-8-Hb-Br/MWCN paper21	MAA	SMX	1.0–100 ng mL–1	–	0.7 ng mL–1	Food samples
MIP-MIL/GCE7	MAA	MEL	10 pM–1.0 μM	DPV	8.21 pM	Liquid milk samples
MIP/COFs-AuNPs/AuE61	o-ATP@AuNPs	AFB1	0.05–75 ng mL–1	QCM	2.8 pg mL–1	Peanut, pistachio, rice, and wheat
PPY/ZIF-67-MIPs/nafion/GCE106	Pyrrole	DA	0.08–100 μM; 100–500 μM	CV, DPV	0.0308 μM	Dopamine injection and human serum
MIP–PB-PC–CNTs/GCE61	Pyrrole	Cys enantiomers	0.1 pM–0.1 μM	DPV, EIS	6.01 fM	Human serum
MIP/UCNPs/CTpBD-Au/GCE61	Thioaniline	DA	10 fM–1.0 μM	CV	2.0 fM	Rat serum sample
MIP/MoS2/NH2-MWCNT@COF/GCE40	pABA	SMR	0.3–200 μM	DPV	0.11 μM	Pork and chicken samples
MIP-Au@NH2-MIL-125(Ti)-Gr/GCE101	l-Cys	BSA	10–18–10–12 g mL–1	CV, DPV	0.415 ag mL–1	Liquid milk sample
MI/MOF50	4-ATP	l-Phe	2.0–600 pM	CV, EIS	0.33 pM	Urine samples
MIP/NGAL103	p-Bis(2,2′-bithien-5-yl)methyl-o-catechol	NGAL	0.1–0.9 μM	CVEIS	0.12 μM	–
MIP/MOF29	PATP-functionalized AuNPs	Gly	1.0 pg L–1–1.0 μg L–1	CVLSV	0.8 pg·L–1	Tap water
MIP-MOF31	PATP-functionalized AuNPs	AFB1	3.2 fM–3.2 μM	CV, LSV	3.0 fM	–
MIP/MOF29	PATP-functionalized AuNPs	GMT	3.8 fM–38 nM	LSV	0.9 pg L–1	Serum samples, drug formulation
MIP/MOF107	PATP-functionalized AuNPs	TNT	4.4 fM–44 nM	LSV	0.04 fM	Tap and natural waters
BC/Cr2O3/Ag/MIP/GCE108	AM and MAA	NFZ	5.0 nM–10 μM	DPV	3.0 nM	Blood, urine, and pharmaceutical agents
MIP/PC/GCE106	Resorcinol	LID	0.2–8000 pM	CV, EIS	67 fM	Blood samples
MOF-MIP109	4-ATP	Thimerosal	0.8–80 pM	CV, EIS	0.035 pM	Chloramphenicol eyedrop samples

^aAbbreviations: MIP: molecularly imprinted polymer; Au: gold; Cu-MOF: AuNPs-functionalized Cu-metal–organic framework; N-GQDs: nitrogen-doped graphene quantum dots; GCE: glassy carbon electrode; CV: cyclic voltammetry; SCCA: squamous cell carcinoma antigen; Ab: antibody; Tyr: Tyrosine; pTH: polythionine; Au@ZIF-67: gold nanoparticles@zeolitic imidazolate framework-67 composite; p-ABA: para-aminobenzoic acid; DFC: diclofenac; Zr: zirconium; GA: gallic acid; Ppy: polypyrrole; MAA: methacrylic acid; ZIF-8: zeolitic imidazolate framework-8; SMX: sulfamethoxazole; Hb: hemoglobin; MWCNTs: multiwalled carbon nanotubes; MEL: melamine; GO@COF: graphene oxide@covalent–organic frameworks; SDZ: sulfadiazine; AP: acetaminophen; PB: Prussian blue; E2:17β-estradiol; CNTs: carbon nanotubes; AFB1: aflatoxin B1; AuE: gold electrode; o-ATP@AuNPs: o-aminothiophenol functionalized AuNPs; RIF: rifampicin; INZ: isoniazid; MC: mesoporous carbon; DA: dopamine; PATP@AuNPs: AuNP-based thioaniline units; UCNPs: upconversion nanoparticles; SMR: sulfamerazine; MoS₂: molybdenum disulfide; NH₂-MWCNT@COF: amino-functionalized carbon nanotubes@covalent organic frameworks; BSA: bovine serum albumin; l-Cys: l-cysteine; Au@NH₂-MIL-125(Ti): AuNPs supported amino-functionalized Ti-benzenedicarboxylate metal–organic frameworks; Gr: graphene; l-Phe: l-phenylalanine; 4-ATP: 4-aminothiophenol; NGAL: lipocalin; TC: tetracycline; Gly: glyphosate; PATP: p-aminothiophenol; GMT: gemcitabine; LSV: linear sweep voltammetry; MMOF: microporous metal–organic frameworks; TNT: 1,3,5-trinitrotoluene; NFZ: nitrofurazone; AM: acrylamide; BC: biomass carbon; Cys: cysteine; PC: porous carbon; LID: idocaine

4. Properties of MOF/COF-MIP Combinations

4.1. Morphological Properties of MOF/COF-MIPs

Morphological studies of MOF/COF-MIPs provide a basic understanding of the inclusion of MIPs along with the MOFs/COFs after the synthesis of the composite. Direct investigation techniques used to obtain a preliminary understanding of a composite’s morphology and roughness are electron microscope characterization, such as scanning electron microscopy (SEM) or transmission electron microscopy (TEM), and scanning probe microscopy (SPM) of MOFs/COFs and MOF/COF-MIPs.

Generally, MOFs/COFs particle size varies from the range of nm to μm; in hybrids, the particle size tends to increase relatively depending on the orientation and size of the imprints. In most cases, MOF/COF-MIP composites retain the morphology of the MOFs/COFs even after the MOF/COF-MIP synthesis (Figure 5).^26,61,79 The surface of hybrids is relatively rough, and the particle size is widened compared with the MOF/COF counterparts, indicating the incorporation of the MIP into the MOFs/COFs.^26,61,79 In some cases, the agglomeration of particles in hybrid composites^21,32 (Figure 5a–d) or core–shell formation (Figure 5e–h)⁵⁸ has been observed after MOF/COF-MIP synthesis. Furthermore, SPM provides insights concerning particle size, roughness, thickness, asymmetry, pore density, interface, and the shape of the cavities.¹⁰⁶Table 3 lists the different combinations of MOF/COF and MIPs that have different morphological characteristics. It compares the critical parameters of MOF/COF-MIPs, such as structure, size, and materials. However, conclusions cannot be drawn concerning morphology based on imaging techniques alone, which requires the assimilation of multiple data sets; thus, combinations of analytical characterizations are necessary. Zeng et al. made use of electron microscopes, FTIR, and adsorption isotherms to understand the core–shell formation and to obtain information concerning the elution of molecules,⁸⁴ while Chen et al. used imprinted polyaniline on ZIF-67 MOFs for the electrochemical detection of tyrosine. The latter authors used not only SEM but also electrochemical analysis to confirm the changes observed through SEM imaging.²¹ In contrast, Chaix et al.¹⁸ used force-based microscopies such as atomic force microscopy (AFM) to investigate the surface of a MOF-MIP (containing 5 nm particles) over a relatively more extensive area of 1 μm × 1 μm. The analysis yielded accurate information on the high phase variation of the hybrid thin film, implying that AuNPs are larger (approximately 30 nm). Therefore, it is essential to adopt a multicharacterization approach for investigation concerning the surface probing of MOF/COF-MIP hybrids.

Figure 5.

Morphology study through electron microscopic imaging: SEM images of (a) UiO-66-NH₂ (MOF) and (b) MOFs/COFs-MIP indicating the change in surface texture and particle widening after MOF/COF-MIP synthesis. Reprinted with permission from ref (32). Copyright 2015, Wiley. SEM images (c) NH₂-MIL-53(Al) (MOF) and (d) MIP/MOF. Reprinted with permission from ref (61). Copyright 2011, Royal Society of Chemistry. TEM images for (e) UIO-66 (Zr) (MOF) and (f) MOF/COF-MIPs. Reprinted with permission from ref (58). Copyright 2020, Springer. TEM images for (g) CTpBD-Au (COF) and (h) MIP/COF indicating the core–shell formation after MOF/COF-MIP synthesis. Reprinted with permission from ref (61). Copyright 2011, Elsevier.

Table 3. Materials Used for MOF/COF-MIP Synthesis and Their Parameters Are Characterized through Imaging.

MOF/COF type	Technique	MOF/COF-MIP configuration	Structure	Size
UiO-66-NH232	SEM/TEM	MIP@UiO-66-NH2	Octahedron	300 nm
Cu-MOF90	SEM	MIP/Au@Cu-MOF/N-GQDs	Crystalline hexagonal	80–20 nm
UMCM-1 MOF33	SEM	MOF-DES/MIPs	Needle-shaped crystalline	∼200 nm
Zr-MOFs51	SEM	MIM/MOFs@G	Regular polyhedron	20–30 nm
CTpBD COFs61	SEM/TEM	MIP/UCNPs/CTpBD-Au	COF-Lamella	∼30 nm
ZIF-67106	SEM/AFM	PPy/ZIF-67-MIP	Rhombic dodecahedral	1.0 μm
MIL-101110	SEM	MIL@MIP	Octahedron morphology	200–300 nm
COF21	SEM	CMIP, MCMIP	Circular	80 nm
ZIF-892	SEM/TEM	C3N4/ZIF-8@MIP	3D dodecahedral	∼100 nm
MIL-10170	SEM	MIL-101@MIP	Spherical	400–600 nm
Zn4O MOF103	SEM	MOF-MIP	Cubic	NA
AuNPs MOF31	AFM/TEM	AFB1MIP-MOF	NA	5.0 nm
TbMOF-7636	SEM	MIP@TbMOF-76	Pillar-like rod crystals	500 nm

4.2. Adsorption Properties

One of the most crucial features of MOF/COF-MIP materials is their capability to adsorb the targeted analyte with high selectivity. Therefore, in this section, we discuss the adsorption properties of the reported MOF/COF-MIPs in detail.

4.2.1. Nitrogen Adsorption–Desorption Analysis

Brunauer–Emmett–Teller (BET) surface area has been widely used to estimate surface area and pore size. These values are expected to be higher in the case of MOFs compared with hybrid counterparts.⁷⁹ Devic et al.⁷⁵ have reported nitrogen BET sorption experiments focused on investigating the surface characteristics of a MOF, MOF@SiO₂, MOF@SiO₂@MIP, and MOF@SiO₂@NIP. The results revealed that the surface area of the MOF decreased after modification with the MIP, confirming the formation of the polymer around the MOF. Moreover, the MOF-MIP exhibited a larger surface area compared with MOF-nonimprinted polymer (NIP), which is due to the presence of MIP cavities created after the removal of the template.⁷⁵ Pan et al. and Parvinizadeh et al. demonstrated the decrease in pore volume and surface area after the modification of MOF with the MIP. They confirmed the successful formation of MIPs and MOFs/COFs composites through various characterization techniques.^79,94 Nitrogen adsorption–desorption analysis can be used to investigate the applicability of MOFs/COFs as supporting materials and sorbents to be modified with the MIP.⁸⁸

4.2.2. Adsorption Characterization

Adsorption characterizations are widely used to explain the adsorption performance of a MOF/COF-MIP toward a specific target analyte in comparison to a MOF/COF-NIP. Indeed, several other studies have been conducted to confirm the adsorption performance, such as adsorption isotherms, adsorption kinetics, selectivity studies, and adsorption thermodynamics. Such studies have been based on the determination of the binding capacities Q according to the following eq 1:

1

where m is the adsorbent weight, V is the volume, C_i is the initial concentration, and C_e is the equilibrium concentration.

The adsorption isotherm represents the equilibrium relation between the binding capacity and concentration at a constant temperature to compare the MOF-MIP and MOF-NIP at different concentrations. Several models have been developed to describe the experimental results, such as the Freundlich, Langmuir, and Scatchard models. When the experimental results fit well with the Freundlich model, the adsorption of the analyte on the MOF-MIP could be considered multimolecular layer adsorption, and the recognition sites were not evenly distributed on the surface of the MOF/COF-MIPs.²¹ The Langmuir model indicated that the binding of the target analyte on the MOF/COF-MIPs had occurred through monolayer adsorption, and the material had the same adsorption energy and affinity to the target analyte all over the surface in the adsorption process.^80,83,87 Mirzajani et al. employed adsorption isotherm data processed by the Scatchard model.³⁵ The developed MOF-MIP had two linear ranges, indicating that this material exhibited two different binding sites. The first indicated the presence of specific recognition sites, while the second was attributed to the presence of nonspecific binding sites. However, the MOF-NIP displayed only one linear range, suggesting that this material had one type of binding site, which was due to nonspecific adsorption. Hence, the adsorption isotherm is a tool for validating the creation of recognition sites in MOF/COF-MIPs from the process of molecular imprinting when MOF/COF-MIPs exhibit higher binding capacities with an increasing concentration of the analyte compared with MOF/COF-NIPs.^{21,35,80,82,83}

Adsorption thermodynamic studies are also essential to investigate the adsorption of the target analyte on the MOF/COF-MIPs at different temperatures. Thermodynamic parameters such as ΔG° (the standard free energy), ΔH° (the standard enthalpy), and ΔS° (the standard entropy) are calculated using the following eqs 2–4:

2

3

4

where K_d represents the distribution coefficient, and C_e and Q_e are the equilibrium concentration of the analyte and the adsorption capacity, respectively.

It has been observed by Zhao et al. that ΔG° < 0, ΔH° > 0, and ΔS° > 0, confirming that the adsorption of the analyte (canidin-3-O-glucoside) on COF-MIP was a spontaneous, endothermic, and random process.²¹ Similar results were obtained with the adsorption of cyhalothrin on the MOF-MIP.³⁵ On occasion, the ΔH° decreases after such concentration of the target analyte, which can be explained by the appearance of physical adsorption between the analyte species.²¹ The adsorption rate of MOF/COF-MIPs can be evaluated by adsorption kinetics; generally, their adsorption rate in the first stage is the fastest before decreasing when more time is allowed.²¹ Two kinetic models, namely pseudo-first-order (PFO) and pseudo-second-order (PSO) can be employed to determine the speed control mechanism. It was reported that the kinetic adsorption of cyanidin-3-O-glucoside on a COF-MIP fits well with the PSO model. The theoretical adsorption amount calculated by this model was closer to the experimental saturated adsorption amount. Thus, the adsorption rate was identified by the square of the number of unoccupied recognition sites on the COF-MIP surface.²¹ The PSO model of MOF-MIP has also indicated that the limiting step of the adsorption rate is mainly the chemical adsorption process.^35,87 The adsorption rate constants determined by the kinetic model exhibited higher values in the case of the MOF-MIP compared with the MOF-NIP, which confirmed that the MOF-MIP exhibited higher kinetic adsorption.⁸⁷ Huang et al. have shown the kinetic difference between traditional MIPs without supporting materials and a MOF-MIP.⁸¹ Their MOF-MIP (MIL-101@MIPs) exhibited fast adsorption compared with the MIP due to its fast mass-transfer rate due to a thin polymer layer around the MOF, which reduced the distance between the adsorption sites and target analyte.

The molecular recognition ability of the MOF-MIP to rebind the target analyte selectively and efficiently in complex samples is a crucial factor that should be studied further. Indeed, several analog components can be selected as competitive substances to perform the selectivity study under similar adsorption conditions.^21,79,87 It can be concluded that the adsorption characterization of the MOF-MIP plays an important role in understanding the properties of these materials. In fact, by comparing the adsorption capacities of MOF-MIPs and MOF-NIPs, it is possible to easily confirm the successful creation of imprinted cavities. MOF-MIPs usually exhibit high adsorption capacity at different temperatures, concentrations, and times. Absorption characterizations are strongly recommended to ensure material selectivity in the presence of other possible interferences before using the MOF-MIP material to detect target analysis in the real samples.

4.3. Electrical Properties of MOF/COF-MIPs

Most existing MOFs/COFs are electrically insulating in their pristine form and used in gas-capture and capacitive-sensing applications.⁶⁵ Researchers have recently attempted to develop strategies to realize semiconducting and conducting MOFs.¹¹¹ The receptor material (MOF/COF) requires specific properties (charge holding/passing element) to transduce a typical analyte’s availability into a quantifiable physical signal. To obtain conductive MIPs, researchers have widely used conducting polymers, such as polypyrrole⁶¹ and polyaniline,⁹⁰ and coupled them with nanomaterials such as AuNPs,^21,61 graphene,¹⁰¹ and MWCNTs.⁴⁰

Moreover, Chen et al.²¹ have recently recognized the poor conductivity of ZIF-67 MOFs and applied the functionalization process with the help of conductive nanomaterials, such as AuNPs, to enhance their conductivity. In addition to AuNPs, other materials are used to improve the conductivity of MOF materials, such as AgNPs,¹¹² CuNPs,¹⁰² PtNPs,²⁷ and conducting polymers.⁶¹ Zeng et al.⁸⁴ utilized a Cu₃(HITP)₂ 2D MOF in a reversible chemi-resistive configuration that was capable of detecting ammonia vapor at subppm levels. The authors made use of copper to catalyze the amine functional groups to imine groups to synthesize 2D Cu₃(HITP)₂. The doping of a MOF-MIP using highly conducting metals has also been reported by Xu et al.⁸⁴ They incorporated PtNPs within the UiO-66 MOF, which significantly enhanced the electrical properties of the MOF/MIP. COFs also suffer from poor conductivity, which limits their use in sensing. However, Manousi et al.⁴⁰ made a composite of GO and COF (GO@COF) to prepare a MIP electrochemical sensor. They achieved high signal amplification for the simultaneous sensing of acetaminophen and sulfadiazine.

To summarize, exploiting MOF/COF-MIPs in sensing requires the enhancement of electrical properties as one of the main features coupled with other vital properties. It is crucial to engineering new synthetic strategies that offer not only high selectivity but also high sensitivity toward analytes, which is only possible when the conductivity of the composite is high.

4.4. Labeling Properties of MOF/COF-MIPs

Recent developments in MOF/COF-MIP combinations have led to the preparation of colorimetric labeled composites using different types of fluorescents and luminescents. Quantum dots (QDs) have recently attracted much attention due to their low cost, biocompatibility, and outstanding fluorescent properties. In addition to QDs, other labeling agents have been used along with the MOF/COF-MIPs (Table 4). This section presents the main labeling techniques and their properties.

Table 4. Features of Fluorescent-Based MOF/COF-MIP Combinations^a.

MOFs/COFs-MIP	Fluorescent/luminescent label	Analyte
MCOFs@MIPs@CDs88	CDs	2,4,6-Trinitrophenol
MIP@MOFs@CdSe/ZnS QDs92	CdSe/ZnSQDs	Pyrraline
CN-grafted COF@MIP51	CNs	4-Ethylguaiacol
CDs@ZIF-8@MIP116	Blue luminescent CDs	Quercetin
MIP@COFs@CdSe/ZnS QDs26	CdSe/ZnS QDs	Quinoxaline-2-carboxylicacid
(MIOP) based on QDs-grafted COFs26	CdSe/ZnS QDs	Tyramine
MIPs/CNs-embedded COFs26	CNs	Tryptamine
CTpBD-Au61	Electrochemiluminescence UCNPs	DA
CuNCs@CuMOF/MIP-Fe3O4@SiO2 NPs117	Chemiluminescence CuNCs@CuMOF	Tramadol
Fe(III)-MIL-88B-NH2200	Electrochemiluminescence ZnSeQDs	Squamous cell carcinoma antigen
QDs-based COFs26	Luminescent QDs-based COFs	Quinoxaline-2-carboxylicacid

^aAbbreviations: MCOFs: magnetic covalent organic frameworks; CDs: carbon dots: MIOP: molecularly imprinted optopolymer: QDs: quantum dots: CNs: carbon nanodots.

4.4.1. Labeling with Carbon Quantum Dots

To date, many strategies based on QD-labeled MOF/COF-MIPs have been developed. Liu et al. proposed a combination of MIP with QDs in the presence of MIL-101.⁹³ The proposed optosensor was based on the MIP/MIL-101 combination and CdSe/ZnS QDs, which makes use of one-step reverse microemulsion polymerization. The composite was used for the detection of pyrraline in milk powders and glycation end products that accumulate in different tissues and organs of the human body.

The use of carbon nanodots and fluorescent NPs provides unique optical properties, such as high biocompatibility, solubility, and electrochemical activity.^113−115 Another optical sensor was developed for the detection of tryptamine (TRY), a biogenic amine, which is a nitrogenous organic compound of low molecular weight in meat samples. Furthermore, a one-pot surface-imprinting synthesis method has been demonstrated. The optosensor consists of MIPs based on carbon dot (CD)-embedded COFs. In this case, MIPs can selectively sense the bonding interactions between COFs and the target molecules and to further transduce these chemical events to detectable fluorescence signals with blue luminescent carbon dots.²⁶

Xu et al.¹¹⁷ have recently documented the first use of ZIF-8 through the combination of CDs and imprinting methods for optosensing of quercetin (QCT). MOFs provide a high specific surface area and porosity, whereas CDs provide high blue luminescence, which act as a signal transducer. The proposed sensor can detect the binding interactions of target molecules and further transduce them to detectable fluorescence signals. Thus, high selectivity and sensitivity can be easily achieved by combining MOFs as an imprinting matrix and quantum carbon nanodot crystals as fluorescent elements.

4.4.2. Labeling with Other Fluorescents/Luminescents

Other labeling agents have been widely used in conjunction with MOF/COF-MIPs. In this context, Dai et al. developed a double-combination sensing platform with a recognition effect and quenching-type electrochemiluminescence sensor (MIECLS) for dopamine (DA) detection (Figure 4A).⁶¹ To increase the ECL of upconversion nanoparticles (UCNPs), AuNPs-loaded COFs were used as carriers and a PATP@AuNPs-cross-linked MIP matrix. A wide detection range (10 fM–1.0 μM) for DA with a low LOD value (2.0 fM) and high recovery rate were obtained in serum samples (93.25–112.97%). In another study, Yousefzadeh et al. performed ultrasensitive and highly selective tramadol detection using the chemiluminescence detection method.¹¹⁸ They encapsulated Cu nanoclusters in 2D Cu-MOF (CuNCs@CuMOF), which enhanced the CL effect of CuNCs@CuMOF on Rhodamine 6G (R6G). To increase the selective detection of tramadol, the extraction was pretreated with a MIP supported on the surface of magnetic Fe₃O₄@SiO₂ nanoparticles (MIP-Fe₃O₄@SiO₂ NPs). The detection range and LOD values were recorded as 0.003–2.5 μM and 0.80 nM, respectively. The labeling techniques based on MOF/COF-MIPs have still not been thoroughly explored, and more research is thus needed in this direction.

5. Molecularly Imprinted MOFs/COFs as a Platform for the Development of Sensors

5.1. MOF/COF-MIPs-Based Solid-Phase Extraction Techniques Coupled to Sensors

Solid-phase extraction (SPE) is a frequently exploited sorbent-based technique in sample preparation for enrichment/preconcentration and the cleanup process. Generally, SPE relies on a solid chromatographic filler material, usually consisting of small porous particles contained in a column device or cartridge. SPE operates in two different modes. The first mode consists of retaining the target analyte on the solid support and not the interference species; the analyte is then eluted with an appropriate solvent. In the second mode, the interferences present in the samples at high concentrations are retained, resulting in a significant cleanup process. The cartridge is conditioned in both modes with an appropriate solvent, and the sample is then passed through the column by gravitational flow or vacuum-/pressure-assisted flow. Although SPE has several advantages, it faces some problems, such as cartridge blocking, difficulty in performing simultaneous extractions, and the fact that it is time-consuming. The dispersive SPE technique can clearly overcome the drawbacks of conventional SPE, primarily when magnetic MIPs are used as sorbents. Several factors play a crucial role regarding the use of MOF/COF-MIPs in SPE. One of these crucial roles is played by the adsorption medium. Indeed, the selection of the optimal adsorption medium depends on the analyte and the holding material. Polarity, solubility, and pH generally are the properties to consider when attempting to achieve increased adsorption efficiency. The Table 5 summarizes the feature of the use of MOF/COF-MIPs as adsorbents for SPE.

Table 5. Summary of MOF/COF-MIPs’ Solid-Phase Extraction Features.

MOFs/COFs-MIP	Monomer-cross-linker	Analyte	Adsorption medium	Adsorption time	Adsorption capacity	Adsorption dosage	Eluent	Regeneration
NH2-MIL-125 (Ti)@DDTC200	MAA-EGDMA	Fluoroquinol-ones	PBS pH 7.0	15 min for river water, 20 min for milk	60.81 mg g–1	2.0 mg mL–1	Methanol-HAc (V/V, 8:2)	At least 6 times
UiO-66@MIP58	AA and MAA-BIS	Tetracycline	Oxalate buffer, pH 4.0	15 min	2200–3000 ng mg–1	0.5 mg mL–1	Methanol	–
MIL-101@MIIP s80	MAA-EGDMA	Zearalenone	Water	–	14.12 mg g–1	–	Acetonitrile/water (9:1 v/v)	Maximum 7 times
GO/ZIF-8 MOF/MIP34	MAA-EGDMA	Sterol and steroid hormones	PBS pH 6.0	30 min	–	–	Methanol	–
MOFs-MIPs35	APTES-TEOS	Pyrethroids residue	PBS pH 7.0	10 min	474.56 mg g–1	dispersant/sample ratio of 3:2	Acetic acid-acetonitrile (5:95, v/v)	–
Ni@MIL-100(Fe)@MIP93	APTES-TEOS	Hydroxychloroquine	pH 9.0	1 min	84 mg g–1	2.3 mg mL–1	Methanol	At least 5 times
MICOFs118	Tp- TAPB	Fenvalerate	n-Hexane	Flow rate of 7.5 mL min–1	0.049 mmol g–1	–	4% Acetic acid in methanol	50 cycles
MIL@MIP23	MAA-EGDMA	Trichlorfon, monocroto-phos	—	20 min	43.6 mg g–1	—	Methanol/water/acetic acid (95:5:2, v/v/v)	–

5.1.1. Adsorption Medium

Li et al. optimized the pH medium to load fluoroquinolones from milk and water sample solutions on MOF/COF-MIPs. The adsorption efficiency of the fluoroquinolones increased as the pH increased up to 7.0 and then decreased. The protonation of amino groups in fluoroquinolones in an acidic medium results in the cationic form of fluoroquinolones, which makes their adsorption weak. In the basic medium, the holding polymer (polymethacrylic acid) and fluoroquinolones are negatively charged, leading to repulsive electrostatic interactions and thus to decreased adsorption capacity. Ma et al.⁵⁸ investigated the effect of pH concerning adsorbing tetracycline on a MOF/MIP from 2.0 to 11. Their results revealed that the enrichment factor reached the maximum values at three different pH values, namely 4.0, 8.0, and 9.0. The nondissociation forms of tetracycline, which exhibit three dissociation constants, were reasoned for this observation. The adsorption capacity of the MOF-MIP increased with the pH up to 7.0 and then decreased in the alkaline medium for pyrethroid residues. This phenomenon was explained by the fact that the affinity of the MOF-MIP toward the analytes was destroyed in an acidic medium, and hydrolysis of the analyte occurred in the alkaline condition.³⁵ Another example was reported by Parvinizadeh et al., who obtained the maximum hydroxychloroquine adsorption efficiency at pH 9.0. The interaction between hydroxychloroquine and the holding material is based on hydrogen bonds. Indeed, the amino groups of APTES on the surface of the MIP form hydrogen bonds with the N–H and O–H groups of hydroxychloroquine. At pH values lower than 9.0, the amino groups are easily protonated, which makes the formation of hydrogen bonds difficult. However, at pH 9.0, which is near the pK_a = 9.76, the formation of hydrogen bonds is favorable.^93,118

5.1.2. Adsorption Dosage (Concentration of Solid/Liquid)

To rebind the maximum amount of analyte, the concentration of MOF/COF-MIPs should be optimized. Li et al. found that 2.0 mg mL^–1 is sufficient for obtaining the maximum recoveries of fluoroquinolones (95.5%–99.0% for milk and 96.5%–99.5% for river water).²⁰⁰ Ma et al. discovered that the amount of the MOF/COF-MIP used to rebind tetracycline had a negligible influence at low analyte concentrations, indicating that the MOF-MIP had high adsorption capacity.⁵⁸

5.1.3. Adsorption Time

The adsorption time is a significant factor in ensuring the maximum adsorption of the analyte on MOF/COF-MIPs. Li et al. studied the effect of ultrasound-assisted adsorption time on the recovery of fluoroquinolones.²⁰⁰ They found that the maximum recoveries were obtained at 15 and 20 min in spiked river water and milk samples, respectively. Adsorption based on the ultrasound technique is faster than the traditional technique because the strong cavitation effect of ultrasonic radiation increases the frequency and speed of molecular motion. Moreover, ultrasound-assisted adsorption uniformly disperses the adsorbent in the adsorption medium, increasing the enrichment rate of the analyte. The authors also reported that 15 min was sufficient to rebind the tetracycline on the MOF-MIP from a chicken muscle sample.²⁰⁰ In another work, 1 min was used as the adsorption time, and no difference was observed through varying this parameter, which explains the abundant adsorption sites on the surface of the MOF-MIP.⁹³

5.1.4. Elution Solution

Following the adsorption process of the analyte on the MOF/COF-MIP, the elution procedure is important to remove the analyte from the imprinted sites. The elution solution is selected based on the analyte’s solubility (it should be highly soluble) without destroying the holding material and imprinted sites. Moreover, the compatibility of the eluent solution with the analytical measurement should be considered. Acetic acid is widely added to destroy hydrogen bonds and electrostatic interactions. Li et al. reported that the rise in acetic acid in methanol increases the recovery percentage up to 20% (v/v) for the elution of fluoroquinolones.²⁰⁰ Similarly, Ma et al. illustrated that the enrichment factor increased when acid was added to organic solvents (methanol or acetonitrile), but the use of acid damaged the MOF structure, and impurity peaks were found in ultraperformance liquid chromatography (UPLC).⁵⁸ Methanol was selected for extraction after several solvents were tested, including methanol, ethanol, acetonitrile, and THF, because it exhibited the highest extraction efficiency. This efficiency is a result of the formation of more effective hydrogen bonding between methanol as a protic solvent and the analyte.⁹³

5.1.5. Regeneration Ability of MOF/COF-MIPs

Li et al. investigated the regeneration of MOF/COF-MIPs by focusing on their adsorption–desorption cycles. Their MOF/COF-MIPs could be regenerated at least six times, which demonstrated good regeneration ability and reusability.²⁰⁰ Huang et al. found that after five cycles of adsorption/desorption, the recovery was stable at 90% and then decreased in the following two cycles but remained higher than 85%. After seven cycles, the recovery decreased to 70% due to the destruction of imprinted sites during regeneration.⁸¹

5.2. MOF/COF-MIP as Chemical Sensors

Owing to the electrical, absorption, and luminescent properties of MOF/COF, various transduction mechanisms have been explored for chemical and biosensing applications.^51,118 Most MOF/COF-MIP sensors rely on the lock-and-key sensing mechanism, which uses synthetic receptors. Prominent sensors using MOF/COF-MIPs that have been demonstrated thus far are based on the absorption and luminescent properties of MOF/COF-MIPs, which are discussed in this subsection. Due to these composites’ excellent electrical properties, they were easily deployed as chemi-capacitors,^72,200 and organic field-effect transistors,^119,120 and practical environmental monitoring has been demonstrated. However, there are no reports on impedance-based sensors using MOF/COF-MIP hybrids. In addition, Iskierko et al. have investigated the MOF-MIP composite, where MOF was used as a sacrificial component to enhance the surface area of the MIP, by demonstrating an extended gate-field effect transistor sensor to sense the protein.¹⁰⁴ As can be seen in Table 6, the analytical performance of various chemical sensors making use of MIP and MOF/COF combinations are presented.

Table 6. Present Analytical Performance of Various Chemical Sensors Making Use of MIP and MOF/COF Combinations.

Sensing materials	LOD	Linear range	Imprinting factor	Analyte	Sensor type	Detection time
CDs@ZIF-8@MIP116	2.9 nM	0–50 μM	NA	Quercetin	Fluorescence intensity	NA
MIPs@MOFs and QDs92	3.9 μM	5.0–1000 μM	62	Pyrraline	Fluorescence intensity	NA
MCOFs@MIPs@CDs88	100 pM	0.3–100,000 nM	8.4	2,4,6-Trinitrophenol	Fluorescence intensity	1 min
MIL-101@SiO2 NPs (MIP)75	13 pM	50–1400 pM	4.3	Japanese encephalitis virus	Fluorescence intensity	20 min
MMIP61	0.018 mg L–1	0.05–1.0 mg L–1	3.17	Zearalenone	Fluorescence	NA
MIP@TbMOF-7636	0.34 ng mL–1	0.8–90 ng mL–1	NA	Cefixime	Fluorescence intensity	3 s
UCNPs/HKUST-1/MIP61	0.062 mg mL–1	0.1–0.6 mg mL–1	1.82	Bovine hemoglobin	Fluorescence intensity	NA
QD-grafted COF-based MIPs15	5.0 μg kg–1/3.0 μg kg–1	0.03–60 mg kg–1/0.02–20 mg kg–1	NA	Ferulic acid	Fluorescence intensity/SPE-MPLC/MS	NA
CN-grafted COF@MIPs51	17 ng mL–1	0.02–1.0 μg mL–1	7.77	4-ethylguaiacol	Fluorescence intensity	15 min
MIPs based on CN-embedded COFs131	7.0 μg kg–1	0.025–0.4 mg kg–1	2.5	Tryptamine	Fluorescence intensity	20 min
MIOP based on QD-grafted COFs26	7.0  μg kg–1, 5.0  μg kg–1	35–35,000 μg kg–1, 20–2000 μg kg–1	5.6	Tyramine	Fluorescence intensity/SPE-HPLC	NA
Fe(III)-MIL-88B-NH2@ZnSeQDs200	31 fg mL–1	0.0001–100 ng mL–1	NA	Squamous-cell carcinoma antigen	Electrochemical luminescence	NA
MIP/COFs-AuNPs/AuE61	2.8 pg mL–1	0.05–75 ng mL–1	3.27	Aflatoxin B1	QCM	NA
MIL-101@MIP110	0.0689 mg L–1	0.1–0.9 mg L–1	NA	Metolcarb	QCM	NA

5.2.1. Fluorescent Sensors Using MOF/COF-MIPs

The intrinsic porosity and luminescence properties of luminescent MOFs allow chemical analytes to be captured and detectable changes in luminescence to be transduced employing the host–guest chemistry. This principle can be adopted to design fluorescent sensors.¹¹⁶ Two strategies are commonly employed in fabricating luminescent MOFs, one of which is synthesizing lanthanide MOFs that can generate fluorescence emission. The other strategy is the encapsulation of luminescent nanoparticles or guest molecules within the framework of the nonluminescent MOFs including luminescent QDs,⁹² CDs, and metal nanoparticles.^113−115 Similarly, to MOFs, COFs also possess excellent luminescent properties. Moreover, the high chemical and thermal stability of COFs makes them a potential candidate for fluorescent sensors.⁵¹ Luminescent COFs generally comprise large π-conjugated building blocks or possess a rigid internal structure with a large π-conjugated system. In addition, the π-conjugated system can be modified by the formation of −C=N–N=C– bonds and −C=N– bonds to enhance the fluorescence intensity and photoluminescence quantum yield.¹²¹

The MIP technique is a versatile method for preparing synthetic receptors by allowing for the presence of molecular recognition sites of target analytes. The imprinted polymer chemistry further allows the addition of fluorophore recognition units to improve the sensitivity and selectivity of the sensing analytes.⁷ Furthermore, the inclusion of MIPs exhibited the intrinsic signal amplification property, and they can be easily fabricated on devices.¹¹⁶

Using semiconducting QDs, carbon nanodots, graphene quantum dots (GQDs), and gold nanoclusters as fluorophores in MOF/COF-MIP hybrid composites have demonstrated a promising enhancement of luminescent properties. QDs can offer high extinction coefficients with a broad extinction range, which makes them superior to alternative luminescent materials. Extensive studies have explored semiconducting QDs such as CdS, CdTe, and CdSe, which can be incorporated into hybrid composites due to their excellent luminescent properties in the visible range.^122,123 Liu et al. demonstrated a novel optical sensor for trace target detection in which MIPs were anchored onto the surface of MOF (MIL-101) with CdSe/ZnS QDs using a one-step reverse microemulsion polymerization technique.⁹³ The developed sensor with MIPs@MOFs and QDs was highly selective and sensitive for the detection of pyrraline in milk powder samples. Pyrraline is an advanced glycation product that accumulates in various tissues of the body, and in the study, the researchers tested real milk powder samples (Figure 6a). The fluorescence intensity of QDs was observed to be quenched when pyrraline bound to the MIPs@MOFs/QDs, which was due to the charge transfer from the QDs to pyrraline. The presence of MIPs in the composite offered selectivity, and the combination of MOFs and QDs offered double signal amplification, which led to higher sensitivity. QD-grafted COFs have also been shown to provide better sensing properties with the inclusion of MIP-based fluorescent sensors. Zhang et al.²⁷ demonstrated QDs@COFs and MIPs sensors for the detection of tyramine in meat samples, especially fermented meat products. Furthermore, the fluorescence intensity of QDs was quenched when the analyte bound to the composite owing to the charge transfer between the QDs and tyramine. The hybrid composites were used as absorbents to simultaneously detect tyramine by optosensing as well as SPE coupled with high-performance liquid chromatography (SPE-HLC), as shown in Figure 6b. The optosensing method is a quick and selective sensing methodology, whereas SPE-HLC was used to obtain high levels of sensitivity, selectivity, and accuracy. Similarly, Wang et al. fabricated MIPs onto QD-grafted COFs for the efficient detection of ferulic acid.¹⁵ Ferulic acid is a natural antioxidant and is commonly found in health products, cosmetics, and medicines. Zhang et al. used a similar methodology involving MIPs@QDs and COFs for the detection of quinoxaline-2-carboxylic acid in meat and feed samples.²⁷

Figure 6.

Overview of the synthesis and sensing approaches of reported chemical sensors (A) MIPs @ MIL-101 and QD sensor preparation to detect pyrraline (PRL). Reprinted with permission from ref (92). Copyright 2019, Elsevier. (B) MIPs@COF and QD sensor preparation to detect tyramine in meat samples. Reprinted with permission from ref (26). Copyright 2016, Elsevier. (C) CDs and ZIF-8@MIPs fluorescence sensor preparation to detect malachite green (MG) in water samples. Reprinted with permission from ref (75). Copyright 2014, Springer-Nature. (D) QCM sensor preparation to sense aflatoxin-B using MIP/COFs-AuNPs. Reprinted with permission from ref (61). Copyright 2011, Elsevier.

Electrochemiluminescence (ECL) is a transducing phenomenon that is widely used for biosensors. The nanomaterial used for these sensors provides the advantages of chemiluminescence and electrochemical control.⁷⁰ Mo et al. developed an ECL-based biosensor using an electro-polymerized dopamine MIP to detect squamous-cell carcinoma antigen (SCCA). The Fe (III)-MIL-88B-NH₂ MOFs with ZnSe QDs acted as an efficient coreaction accelerator with which the authors could achieve three-way signal amplification for ZnSe QD emission, and the MIPs allowed the selective detection of SCCA. The sensing methodology provides a rapid and selective detection technique for SCCA and has significant potential in terms of clinical applications.⁵⁸

CDs are another class of luminescent nanoparticles, and they possess considerable advantages over QDs in terms of photoluminescence stability, low toxicity, good water solubility, and excellent biocompatibility.¹¹⁶ Xu et al. synthesized CD-embedded MOFs/MIPs for the first time using a one-pot surface imprinting process for the highly sensitive and selective detection of quercetin (QCT).¹¹⁷ In this work, MIPs on the surface of ZIF-8 MOF were designed for specific adsorption of target molecules through recognition sites. Embedded blue luminescent CDs tended to sense the selective binding events and transduce them into detectable changes in fluorescence intensity. The developed CDs@MOF@MIP composite exhibited short adsorption equilibrium, fluorescence stability, high sensitivity, and selectivity toward QCT. Recently, Liu et al.⁷⁶ demonstrated also a fluorescence sensor using CDs and ZIF-8@MIPs with a core–shell structure, prepared using a sol–gel method for selective malachite green (MG) in the water samples (Figure 6c). MG is a synthetic dye commonly used in fishing industries; excess MG is harmful to fish and even toxic to humans. The fabricated sensor provides rapid detection, enables quantitative monitoring of MG, and offers an inexpensive method compared with existing analytical techniques. CDs have also been embedded as fluorescent nanoparticles in COFs-MIP composites. Zhang et al.²⁷ have shown the rapid detection of tryptamine using CDs@COFs/MIPs synthesized using a one-pot surface imprinting technique. This optosensor works based on tryptamine–carboxyl interaction, which results in the quenching of fluorescence through the charge transfer. Liu et al. also used a strategy that involves embedding CDs in COF/MIPs to detect the aroma compound 4-ethylguaiacol (4-EG) in wine samples.⁵² Recently, 2,4,6-trinitrophenol (TNP), which is used in explosives and fungicides, has also been shown to be detectable using a fluorescent sensor based on a composite of magnetic COFs, CDs, and MIPs.⁸⁸

UCNPs are emerging fluorescent nanoparticles with low toxicity, long lifetimes, and a lack of autofluorescence compared with the traditional luminescent materials discussed previously. Hence, UCNPs are widely used for biological imaging applications. Moreover, they can convert NIR light into UV–visible light by multiple photon absorptions.⁶¹ Guo et al. made use of UCNPs and prepared a composite with HKUST-1 MOF. They used a thermosensitive imprinted material that changes its physical properties in response to the temperature. The authors employed this novel strategy for enriching and sensing the protein. The thermosensitive imprinted polymer benefited from the specific recognition of the template protein, which could be controlled by the external environment temperature.⁶¹

Beyond the addition of luminescent materials for enhancing sensitivity, other strategies exist for improving the sensing performance of fluorescent MOF/COF-MIP sensors. One example is embedding preconcentrators in the luminescent MOFs. Eskandari et al.³⁷ demonstrated a fluorescent sensor with a MOF-76-encapsulated MIP, with GO as a preconcentrator and a selective absorbent to detect cefixime antibiotic. Yang et al. illustrated the metal chelation and passivation technique to enhance the selectivity of a MOF-MIP sensor toward the Japanese encephalitis virus.⁷⁵

5.2.2. QCM Using MOF/COF-MIPs

With adsorption of the target analyte being a key feature of MIP, the quartz crystal microbalance (QCM) is another promising transducer for sensing applications of MOF/COF-MIPs. A QCM is an electromechanical sensor, and its resonant frequency is highly sensitive to the mass of the crystal, where the adsorbent is usually placed for sensing. The change in mass due to the adsorption of the analyte results in the shift of the resonant frequency.¹⁶ Qian et al.¹¹¹ employed a QCM sensor to detect an insecticide analyte (metolcarb) using MIL-101 MOF and MIP composite. They surfaced MIP nanoparticles onto the MIL-101 MOF, and the resultant core–shell architecture facilitated excellent adsorption properties. Yao et al.⁸³ adopted a similar strategy in synthesizing UiO-66 MOF-MIP to detect tyramine using QCM while Gu et al.⁶¹ demonstrated an aflatoxin sensor using an AuNPs-doped COF-MIP composite. The addition of AuNPs resulted in additional recognition sites in the synthesized composite. Figure 6d provides an overview of the sensor preparation. All these studies suggest that QCM is an alternate technique that could be explored in the absence of luminescent MOF/COF-MIPs and when attempting to exploit of the adsorption properties of the composite.

5.3. Electrochemical Sensors Based on MOF/COF-MIPs

5.3.1. Food Safety

Electrochemical sensors based on MOF-MIPs have drawn significant attention in the field of food safety owing to their low cost, high sensitivity, and ease of analysis with minimal sample pretreatment. They have been widely used in quality detection in many food samples, such as meat, milk, honey, veterinary drugs, and feed additives.¹¹⁶ Bougrini et al. developed the first MOF-MIP electrochemical sensor with specific recognition sites for tetracycline detection, which is an antibiotic commonly used to treat several infections. The functionalization was prepared through electropolymerization of the aniline monomer and subsequent modification with AuNPs in the presence of tetracycline as a template molecule. This approach exhibited a significant increase in the detection limit of tetracycline, even in complex matrices, such as that of honey.³¹ Another synthetic antibacterial agent, sulfathiazole (STZ), an important member of the sulfonamide group, was detected by a GCE-based electrochemical sensor prepared as a polypyrrole MIP (imprinted with STZ) and harnessed the signal probe of CuS. The sensor exhibits high selectivity and high electrical conductivity of AuNPs@COF to enhance the electrochemical signal. This study was the first to employ CuS as an electroactive probe for the construction of an electrochemical sensor for detecting STZ. Findings guided the development of improved electrochemical sensors for the ultrasensitive detection of sulfonamide residues in meats, fodder, and the environment.⁸⁸

Another study has reported a biomimetic sensor modified with ferriferrous oxide-reduced graphene oxide (RGO) nanocomposites for the detection of methamidophos or omethoate, a highly active and residual organophosphate insecticide in vegetables. The developed sensor had high sensitivity, good reproducibility, and a wide linear range toward methamidophos and omethoate insecticides. These insecticides have been banned from agricultural use in many countries due to their acute toxicity. This method can be a promising detection tool for the analysis of multiple pesticide residues in vegetables.¹⁶

Fruit toxins such as patulin have also been determined using MOF-MIP-based electrochemical sensors. Hatamluyi et al. introduced an innovative MOF-MIP-based sensor for the ultrasensitive and selective detection of patulin, which is a toxic secondary metabolite produced by certain species of fungi. The sensor is based on a GCE decorated by nitrogen-doped GQDs (N-GQDs) and AuNPs-functionalized Cu-MOFs. The prepared electrode surface is then decorated with a layer of MIP by electropolymerization. The high accuracy and precision of the sensing system in apple juice samples proved its strong potential for the rapid and low-cost determination of patulin compared with chromatographic methods.⁹⁰

Phenolic compounds have recently been identified as potential analytes for MIP-MOF-based electrochemical sensors. Gallic acid (GA) is one of the main phenolic components that occur naturally in plants (including gallnut, blueberries, tea leaves, and black tea). GA possesses antioxidant, anti-inflammatory, antimicrobial, and radical scavenging properties; it also exhibits DNA polymerase activity and ribonucleotide reductase activity. GA is commonly used in food and cosmetics because it can prevent lipid peroxide accumulation and rancidity. A new sensing platform was constructed by combining molecular imprinting technology and mesoporous materials for the highly sensitive detection of GA. The MIP/Fe₃O₄@ZIF-67/Au electrochemical sensor realized the advantages of a short time of analysis, simple operation, strong specificity, high sensitivity, and low LOD for GA. The developed sensor was also successfully used to determine the trace GA in black tea, green tea, and water samples.⁷¹

5.3.2. Pharmaceutical Analysis and Environmental Monitoring

MOF/COF-MIP composite-based sensing approaches have proven to be efficient methods for different applications in environmental monitoring as well as in pharmaceutical analysis. In this context, Guo et al.¹⁰⁸ developed a MIP/MOF-based electrochemical sensor for the detection of TNT, an explosive agent, in tap and natural water samples. Their sensor was found to successfully detect TNT at a very low concentration range (4.4 fM–44 nM) with a low LOD (0.04 fM). Another strategy for the detection of estradiol, an estrogen steroid hormone, was demonstrated in water samples. The developed MIL-53/MIP-based sensor exhibited high sensitivity (LOD 6.19 fM), a wide linear range (10 fM–1.0 nM), and excellent selectivity.⁶¹

Recently, MOF-MIP sensors have also been used for pharmaceutical analysis. Nitrofurazone (NFZ), a commonly used antibiotic for skin infection and trypanosomiasis treatments, was used as an analyte in a study combining a MIP electrochemical sensor with a biomass carbon (BC)/Cr₂O₃/Ag/MIP coating on a conductive AgNPs surface. The resulting imprinted sensor exhibited high sensitivity, a wide linear range, and a low detection limit of 3 nM. Moreover, the developed sensor exhibited high selectivity in the presence of possible interfering species such as furaltadone hydrochloride, semicarbazide hydrochloride, glucose, ascorbic acid, and dopamine hydrochloride, sucrose, starch, and lactic acid samples.¹⁰⁸ A recent study has reported the use of a Zr-MOF-MIP-based electrochemical sensor for diclofenac (DFC), an anti-inflammatory drug used to treat pain and inflammatory diseases. The sensor was easily prepared through electropolymerization and exhibited excellent analytical features toward the detection of DCF in pharmaceutical formulations with satisfactory recovery.¹⁰⁵

5.3.3. Clinical Diagnosis

The design and development of electrochemical sensors based on MOF/COF-MIPs have gained increasing attention in clinical diagnostic applications. Tang et al. described a Fe₃O₄@ZIF-8-MIP-based electrochemical sensor platform that was designed for the detection of sarcosin (SAR), a prostate cancer biomarker.⁸⁰ The analytical performance of the sensor platform was determined by CV measurements, and the corresponding linear range and LOD values were found to be 1.0–100 pM and 0.4 pM, respectively. In the sample application study, the standard addition method was performed for healthy human and cancer patient urine samples. The results revealed that the developed sensor was selective, sensitive, and reproducible for real case studies.

In another study on cancer biomarker detection, Mo et al. developed a rapid and sensitive electrochemiluminescence sensor platform using Fe(III)-MIL-88B-NH₂@ZnSeQDs/Ab as a signal probe for SCAA detection.⁵⁸ In this study, DA was used as the functional monomer for the electropolymerization of MIP film to increase the sensor’s selectivity. The obtained linear range for SCAA was 0.001–100 ng mL^–1, and the LOD was 31 fg mL^–1. The real sample analysis of SCAA was successfully demonstrated in a spiked human serum sample. Furthermore, Rawool et al. designed a sensitive and selective electrochemical sensor using Cu-MOF/mesoporous carbon (MC) for the simultaneous detection of rifampicin and isoniazid.¹⁰³ The electrochemical behavior of the sensor was investigated using the DPV, CV, AdSDPV, and EIS methods. The linear range was 0.08–85 μM for both RIZ and INZ, and the LOD values were 0.28 nM (RIZ) and 0.37 nM (INZ). Simultaneous quantifications of RIF and INZ in blood serum and urine samples were successfully performed. In another study, Chen et al. designed a MIP/pTH/Au@ZIF-67/GCE electrochemical sensor based on the dual-signal strategy for tyrosine detection in human serum samples.²¹ The sensor preparation started by decorating the GCE surface with Au@ZIF-67 and continued with the electropolymerization of polythionine (pTH) polymer on the Au@ZIF-67 layer. The MIP surface was prepared by the CV method using aniline as the functional monomer and tyrosine as the template molecule. The LOD of the sensor was 0.79 nM, and the linear range was between 0.01 μM and 4 μM.

For the detection of gemcitabine, an anticancer drug, Florea et al. created a MIP MOF-based electrochemical sensor.³⁰ Electrochemical characterization of the electrodes was performed by linear sweep voltammetry (LSV); the linear range was found to be between 3.8 fM and 38 nM, and the LOD was 3 fM. Moreover, Duan et al. have been able to detect cysteamine enantiomers by designing a sensitive and selective electrochemical sensor based on Prussian blue (PB)-porous carbon (PC)-CNTs.⁶² Cysteine was used as a template molecule, and PPy was used as a functional molecule for the film of MIP. The characterization and analytical performance of the sensor were performed with DPV, CV, and EIS methods. Cys detection was successfully performed in human serum samples, and the linear range for l-/d-cysteine was found to be 0.1 pM–0.1 μM. Wu et al. described the design of an amperometric sensor platform for detecting l-phenylalanine using MI-MOF film and β-cyclodextrin-functionalized AuNPs.¹²⁴ Analytical characterization of the sensor was performed using CV and EIS electrochemical methods, resulting in a linear range and LOD of 2.0–600 pM and 0.33 pM, respectively. In addition, Zhang et al. developed a MIP-based DA sensor in DA injections and human serum samples.¹⁰⁷ The GCE was modified with PPy/ZIF-67/nafion hybrid, and two linear ranges (0.08–100 μM and 100–500 μM) were found for DA measurement with DPV. In another study, Lu et al. developed a quenching-type electrochemiluminescence sensor for the detection of DA, using CTpBD-Au as the COF material.⁶² The designed sensor platform was found to have a wide linear range and low LOD in the rat serum sample, leading to promising results and the potential for future applications in other clinical fields. In addition to neurotransmitters, Zhang et al. developed a method for detecting lidocaine by using the MIP-based electrochemical sensor platform in blood samples.¹⁰⁷ An isoreticular MOF-8 (IRMOF-8) was coated onto a GCE, and a MIP film was then produced using the electropolymerization of resorcinol in the presence of LID. The sensor exhibited a linear response to LID in the range of 0.2 pM–8 nM. The obtained detection limit was 67 fM.

6. Current Challenges and Outlook

The molecular imprinting combination with MOFs and COFs has proven to be particularly successful, and great progress has been done in the past few years. However, some challenges exist concerning individual processes conducted for MIPs and MOF/COF composites. For example, the process of elution in MIPs (whereby the MIP is obtained from MIP adduct) affects both MOFs and COFs morphologically and prompts fundamental changes. By contrast, the process of MOF/COF activation with solvents and temperature can affect the intrinsic MIP properties. Therefore, the focus should be on the potential future scope with perspectives based on attributes such as synthesis compatibility (MOF/COF and MIP), applications, market, and environmental friendliness.

It is worth mentioning that although MOF/COF-MIP combinations have been widely explored in different applications, their acute compatibility with a particular transducer or transduction mechanism cannot be generalized, and it must be determined for every procedure. Another issue that should be considered is that MOFs/COFs porosity/sorption properties can also be altered with the increasing or decreasing of temperature. This gives MOFs/COFs an edge over other materials and retains a higher volume of a target analyte in bulk and/or passes it on to the underlying MIPs. Similarly, a MIP layer on top of MOF/COF can lead to alterations to their physical properties, thereby establishing a unique/selective detection process. However, such heterostructure combinations might be meaningful only when there is a requirement for analyte classification, preconcentration, secondary transduction material, and charge exchange.

All the above-mentioned criteria are directly linked to the geometrical aspect of the transducer and process approach. In the latter case, researchers have also taken the approach of core–shell structures,^70,124 with the MOF/COF being the core and the MIP being an external shell. Most reports on sensor applications on MOF/COF-MIPs to date have adopted core–shell structures due to their ease of processing.¹²⁴ However, the additional step of preserving the MOF/COF structure under reduced pressure is always essential before proceeding to the incorporation of a MIP. An additional advantage of this configuration is the isolation of MIP cites from the MOF/COF processes, such as solvent activation, high reaction temperatures (for crystallinity), and long reaction time. The opposite can occur during the process of elution in MIPs, wherein the corresponding solvent can infiltrate into the pores of MOFs/COFs, eventually damaging them. Such irregularities can be avoided to some extent by employing orthogonal solvents. For instance, an orthogonal solvent (for SBUs in MOFs/COFs or MOFs/COFs overall) can be used as elution solvents in MIPs, and, contrariwise, activation solvents in MOFs/COFs can be orthogonal to the MIP elution process. MOF-MIP materials, in contrast, can also cause an increase in the sensitivity of the overall sensor device. This increased sensitivity essentially requires the MIP to be bedded tightly in the MOF while also conformally covering its surface.

There is still a significant need to use ultrasound-assisted synthesis techniques for the preparation of MOF/COF-MIP combinations. Indeed, their use will afford many advantages, including the acceleration of the polymerization reaction rate, the generation of homogeneous polymer-chain growth, higher yields, and milder conditions. In addition, they have the potential to assist the formation of more desired spherical nanoparticles.^3,55 Moreover, ultrasound-assisted techniques will likely allow for significant savings in terms of energy and time.

Delving into flexible electronic applications, where sensors are often considered essential components, MOF/COF-MIP combinations can excel. MOF/COF-MIP shells can be advantageous here, as the polymer matrix of MIP can be made to conformally deposit on flexible platforms or the polymer matrix can be made to bend up to a certain angle. However, such a composite can also cause problems such as the expansion of the template framework and the loss of ideal characteristics. Hence, such a requirement calls for the functional groups in template cavities to strongly adhere to some chemical bonding with the analyte molecule. This approach can help transfer the binding effects of chemisorption into electrical signatures in a wearable platform. Gas-sensing applications are considered a primary requisite and can satisfy the current enormous demand for sensors, ranging from smart cities to oil rigs.

To date, only one work has reported the use of MOFs or MIPs in a single platform for the detection of gases.¹²⁵ Schütze et al. successfully utilized a solid-state sensor platform to realize a sensor array for indoor air-quality assessment. The demonstrated system was capable of preconcentrating and quantifying gas molecules, but further optimization is still required from two different perspectives. (i) Unlike the chemiresistance used in the system, which requires mass spectroscopy, which is bulky and complex. (ii) The receptor material should be upgraded to a MOF/COF-MIP composite and not just either MOF or MIP. As discussed previously, both MIPs and MOFs/COFs are individually credited for their sensing capabilities and are also considered to demonstrate commendable performance in synergy. Thus, a viable option is to make use of this synergetic combination for gas-sensing applications with less complex transducers.^126−130

The outdoor usage of plug-and-play devices in sensor systems, such as in hand-held prototypes, in humid and hot or cold climates, calls for frequent device calibrations. Hence, sensor units require a receptor material that is durable, stable, reusable, and user-friendly. MOF/COF-MIPs’ synergistic combination can achieve those qualities due to the diversified individual properties and may thus become coveted materials for the sensor market. Commercially, both MIPs and MOFs/COFs have proven highly successful, a fact reflected in the recent achievements of MIP diagnostics limited to the area of sensors. Fluorescent Nano-MIPs have been introduced onto the market for the detection of target analytes related to drug abuse, disease, and pharma. Similarly, in the case of MOFs/COFs, MOF Technologies (Canada) and Nu-mat Technologies (USA) have carved a niche for themselves in the chemistry and technology domains. They stand at the forefront in terms of not only making MOFs/COFs a commercial reality but also in pioneering their production. Applications such as gas separation, storage, delivery, and biodiagnostics have been the primary focus of both companies while they concomitantly scale up their production. We posit that the inclusion of MIPs in these tiny entities will help enhance the intrinsic properties of the hybrid structure such as stability, control, and functionality.

7. Conclusions

MOF/COF-MIP-based sensors have gained increasing attention recently. This review covered the different synthesis methods for MOF/COF-MIP combinations using electrochemical and chemical synthesis methods. A significant number of MOF/COF-MIP sensors have been proposed in the past decade and successfully applied for the detection of different analytes of environmental, clinical, biomedical, and food interests. It can be noted that novel combinations of these hybrid materials have been configured with different kinds of transducers, including optical, electrochemical, and mass transducers. Thus, more sensitive and selective sensors have been developed for different applications. Furthermore, MOF/COF-MIP combination-based sensors have contributed to the enhancement of mechanical/thermal stability, provided high adsorption capabilities, and exhibited a large surface area.

The developed MOF/COF-MIPs have been widely used for chemiluminescence and fluorescence applications as well as for SPE applications, due to their outstanding features. Electro-synthesized MOF/COF-MIPs on top of electrochemical transducers have also been announced recently. These conductive substrates allow the achievement of high sensitivity and selectivity of the desired analyte, and their imprinted layer thickness can be easily controlled. Thus far, the results reported in this field are promising and indicate the possible innovation of low-cost, sensitive, selective, and miniaturized sensors. However, more research is still needed to develop MOF/COF-MIPs sensors with long-term stability, reusability, and low costs, which will likely find more applications in the market.

Author Contributions

^⊥ These authors contributed equally to this work.

The authors declare no competing financial interest.