Deposition Methods for the Integration of Molecularly Imprinted Polymers (MIPs) in Sensor Applications

Offering high specificity and selectivity, molecularly imprinted polymers (MIPs) are synthetic polymeric affinity reagents that have become increasingly popular over the last couple of decades. Due to their long‐term chemical and physical stability and low production cost, they have become an increasingly popular choice of receptor in the realm of sense. MIPs have therefore been associated with the detection of small molecules, proteins, cells, and pathogens, proving a highly robust and useful tool in the production of next‐gen sensing platforms. This said, the development of these sensors pivots on one simple fact; these receptors have to be deposited onto a substrate for their desired application. The deposition of MIPs during sensor fabrication is therefore of great importance, with the field utilizing an array of mechanical and chemical deposition methods to achieve this. To this end, this review, therefore, sets aim at coalescing these different deposition approaches, classifying them, and outlining their utility when it comes to receptor design and integration. Thus, offering a knowledge base on current deposition methods, potential future approaches and analyzing where the MIP deposition field is tending toward.


Introduction
The concept of molecular imprinting was introduced as early as the 1930s, while first records of molecularly imprinted polymers (MIPs) date back to the 1980s with pioneering work by Mosbach, Sellergren, and Wulff, synthesizing MIPs based on noncovalent and covalent interactions, respectively. [1][2][3][4] MIPs are generally defined as highly cross-linked synthetic polymeric structures that are able to selectively bind target analytes, thus DOI: 10.1002/adsr.202200059 mimicking the natural antibody-antigen mechanism. [5] The presence of highly specific nanocavities within the polymeric matrix compliment the morphology and chemical functionality of the desired target species, facilitate this "lock and key" interaction that is associated with biological receptors. Critical constituent components (e.g., functional monomer, functional crosslinker, porogenic solvent, and initiator) are used to achieve this, forming a complex interwoven structure of ionic interactions, hydrogen bonding, and Van Der Waals interactions between the template species and these components prior to polymerization. [6] Once initiated under thermal or photochemical conditions these components form a rigid polymeric structure encapsulating the target species, capturing its likeness and generating said nanocavities. After polymerization, the target species is extracted by mechanical means or by washing with appropriate solvents under Soxhlet conditions, yielding a vacant polymeric material ready for use (Figure 1). [5] Early research focused on the development of these receptors towards small molecules, though as the field has expanded so has the list of potential targets with proteins, bacteria, fungi, and viruses being key species of interest. [7][8][9] In turn, this catalyzed the development and utilization of new polymerization methods that could facilitate the synthesis of MIPs toward more complex analytes. [10][11][12] These synthetic approaches can be divided into three main categories, polymerization in solution, solid phase synthesis, and electropolymerization. [12][13][14][15] Methods such as bulk, precipitation, and suspension polymerization fall under solution-based synthesis, following the traditional route of synthesis already expressed. [16][17][18] The primary drawback of these methods lies in poor homogeneity in particle shapes and sizes, which is reflected in the reproducible performance of the receptors. [19] This said, these methods offer the advantage of straightforward synthesis and therefore remain a staple for the synthesis of MIPs for small molecules. [16,17,20] In an attempt to mitigate these negative aspects, a surface grafting approach utilizes solution based synthesis by facilitating receptor layer formation directly onto a substrate by means of covalent bonding. [21] Alternatively, solid phase synthesis utilizes the immobilization of the template species onto a solid support, eliminating the need for the harsh extraction  methodologies. [12,22,23] Instead opting for an approach that removes low-affinity receptors by washing the solid support with cold water, and the higher affinity MIPs with hot water. Thus yielding MIPs with a higher degree of homogeneity in terms of template affinity, but also in relation to shape and size. [12] As the template species is bound to a solid support, a whole host of templates are possible (e.g., molecules, peptides, and proteins) with the main drawback of the approach being associated with the low yields of the collected high affinity MIPs. In contrast to these categories of polymerization, electropolymerization enables the polymerization of a conjugated monomer directly onto a substrate and yields a polymeric film instead of particle. [11] Though this draws parallels with a surface grafting, the deposited layer is noncovalently bound to the surface and relies upon adsorption to the surface instead. [24] The substrate functions as the working electrode and the polymer film is deposited directly by anodic oxidation. [25] One of the major advantages is that this enables precise control over the formed polymer film thickness by carefully choosing the reaction conditions, increasing the reproducibility of the imprinting process. [26,27] Overall, the various approaches allow the production to be tailored towards a desired template species, with the resulting receptors having excellent physical properties (resistance to harsh temperatures and pH) that can be applied in a variety of scenarios (sample pre-treatment, chromatography, targeted drug delivery, purification, and sensing. [5,[28][29][30][31][32][33][34][35] Of these applications, coupling MIPs with sensing technologies has become ever more popular over the last decade, with a large proportion of MIP based research being focused at the development of MIP based sensing platforms. [29,36] Integration of MIPs into sensing platforms is achieved by coupling the receptor with a transducer element that can translate a binding event at the surface of the MIP into a tangible signal. The technologies that are currently associated with MIPs can be divided into electrochemical, optical, mass-sensitive and thermal devices (Figure 2). [37][38][39] Electrochemical sensing can employ a variety of electroanalytical techniques such as voltammetric, amperometric, or impedimetric measurements, which can be categorized by whether the response is generated by the analyte binding to the receptor or by the analyte itself. [36] Amperometry is an example of the former, wherein the setup consists of two electrodes being separated by a MIP-membrane. Upon analyte binding, the polymer undergoes conformational changes, affecting counter ion diffusion, resulting in a measurable change of the membrane electroconductivity. [40] An example of the latter are voltammetric measurements, where an electrochemically active analyte is either oxidized or reduced on a MIP-modified electrode. [41] The modification of the electrode with MIPs allows for selective adsorption of the analyte and prevents interfering species from reaching the electrode. Furthermore, selectivity can be further increased by employing a suitable potential range allowing for a more precise determination of the analyte due to its oxidation or reduction at a characteristic potential. [42] Optical sensing, on the other hand, relies on the measurement of changes in optical properties, like fluorescence or surface plasmon resonance (SPR) upon binding of the analyte and the following transduction into an electronic signal. [37] The most commonly employed optical property for sensing is fluorescence because of its high sensitivity resulting in low limits of detections (LODs). [43] Prominent examples are core-shell materials consisting of a fluorescent quantum or carbon dot core and a MIP-shell as the recognition element, in which fluorescence quenching is observed upon analyte binding. [44,45] Even lower LODs can be achieved by using SPR as the measured optical property. [46] Similarly to fluorescence sensors, SPR sensors are often core-shell materials, with the core usually consisting of metallic nanoparticles like gold NP that are sensitive toward changes in refractive index upon rebinding. [47] Another extremely sensitive readout approach is the detection of surface mass loading in, e.g., quartz crystal microbalances (QCM). This technique makes use of the reverse piezoelectric effect of quartz crystals, wherein a mechanical wave is brought to resonance by administering an AC-electric field. Mass loading on the surface of the crystals will lead to a change in its resonance frequency which is monitored over time. [48] The advantage of using QCM technique is the ability to detect a large amount of different analytes with little consideration of other physical properties, which is shown by piezoelectric sensors designed toward molecules, proteins and bacteria. [49][50][51] In spite of this, the mass sensitivity is limited when low molecular weight targets want to be detected.
Calorimetric methods utilizing MIPs also exist, though recently a straightforward, versatile method has emerged in the field of thermal based detection known as the "heat-transfermethod" (HTM), which proved its efficiency in the sensing of small molecules, proteins, and bacteria. [7,20,39] Herein, MIP particles or imprinted polymeric layers are deposited onto a solid substrate, which is in contact with a liquid phase. The solid phase is heated to a constant temperature by a heat source and simultaneously, the temperature of the liquid phase is monitored. The heat flow travels through the MIP coated recognition layer and upon analyte rebinding a change of thermal resistance can be observed by a change in the liquid's phase temperature. [39] Though all these technologies incorporate and utilize MIPs in completely different ways, they all require these synthetic receptors to be deposited onto a transducer to function. This is a crucial step in the sensor fabrication, and is often overlooked or neglected. To date, there is no one defined method of coupling a MIP with a specific transducer, with researchers currently using a range of different techniques to do so. This review therefore wishes to chronicle and differentiate the formats in which MIPs can be deposited. Thus, highlighting each approach and potential modifications that can be conducted to integrate synthetic receptors for specific readout technologies.

Mechanical MIP Deposition
The integration of MIPs into sensing platforms can be achieved through several approaches; the choice of a specific approach over another must be carefully considered as it affects the readout technology used and thus the potential applications of the sensor. [52,53] Several mechanical deposition methods have been used in recent years to produce MIP-based sensor platforms capable of targeting a wide variety of targets, from small molecules and proteins to viruses and bacteria. [54] In these methods, the sensor production process requires little or no chemical functionalization, making these methods very attractive in terms of marketability, which is considered a major challenge for MIP-based sensors. [55] In this chapter, we will focus on approaches achieved via physical deposition methods, such as microcontact stamping, drop casting of MIP particles or of the pre-polymerization mixture, and the incorporation of MIPs into electrodes.

Microcontact Stamping
Most commonly, microcontact stamping finds its home in the world of soft lithography where it is used to shape polymeric surfaces by introducing a mold. [56] The value of this process has therefore been realized by the MIP field, with its primary application in the imprinting of macromolecular targets. [57][58][59][60] In this approach, two solid substrates are prepared and then placed on top of each other in a "sandwich" conformation; one of the substrates presents the template immobilized on a stamp and the other one represents the solid support, usually functionalized with a viscous oligomer layer. While the two substrates are in contact, the crosslinking process is completed and the stamp with the immobilized template is removed leaving cavities in the formed polymeric layer that resemble the target in shape and functionality [61] (Figure 3).
A variant to this approach has also been used for the deposition of MIP microparticles on a solid support. In this method, polymer particles are immobilized on a stamp and then this is applied to an adhesive layer, resulting in the formation of a MIP layer. The stamp used to immobilize the particles is usually PDMS, and as for the adhesive layer different polymers, such as PVC or conjugated polymers (e.g., PPV), have been used successfully. The   formed receptor layer is then used as sensing element for different types of transducers, such as HTM [20,[77][78][79] or EIS [80][81][82][83] allowing the detection of several markers. The major drawback of this approach lies in the irreproducible nature of the formed layers, primarily stemming from how the analyte species is stamped into the polymeric layer. To date there has been no in-depth study that verifies how the template/MIP is distributed onto the polymer layer during the stamping/curing, making the exact replication of functionalized substrates near an impossibility. This critical factor must be addressed if the technology is to be mass-produced, otherwise each stamp will afford a different surface coverage and particle/imprint distributions therefore lending itself to heterogeneous and irreproducible receptors.
Some progress has been made towards providing a solution to this conundrum, with concepts such as producing a "master stamp" showing great promise. This concept is exemplified by Seidler et al. where they introduce the idea for the imprinting of yeast. [84] To this end, PDMS is poured over a yeast imprinted surface that was synthesized prior and left to cure for two days, thus enabling a stamp to be generated that captures the likeness of the imprint sites but in silicone instead. After the curing process, the PDMS stamp is removed, and can be subsequently used for the production of further functionalized surfaces. The main idea is that the "master stamp" can be used to produce imprinted layers with the exact same distribution of receptors sites and surface coverage as the original layer, and therefore increases the reproducibility and standardization of the layers formed. Das et al. take this concept one step further by combining this process with continuous roll-to-roll imprinting enabling the mass manufacture of imprints for blood cancer cells. [85] Though promising, these modified versions of microcontact imprinting still have their limitations. As the "master stamp" is used more, the resolution and definition of the stamp reduces meaning that the quality of the imprints also declines. This is not an issue in the short term but for mass production this becomes problematic, as more master stamps have to be produced and the process begins to draw parallels with the original microcontact imprint methodology.

Drop Casting
A simple and rapid technique for immobilizing micro-and nanoparticles on different substrates is the so-called "drop casting". This deposition method, as the name suggest, relies on the formation of a layer on a solid flat surface by adding a drop of suspended particles and leaving it to "stand & dry" in order to evaporate the remaining solvent. In the case of MIP particles, this technique is mostly used to immobilize the particles onto different types of electrodes. Different works have reported the drop casting of suspended MIP particles on sensing platforms without addition of supplementary components. [86][87][88][89][90] However, lack of control over particle distribution and limited adhesion at the surface leads to significant limitations in terms of reproducibility. In order to enhance this process, the technique can be modified by mixing the suspended imprinted particles with other components used as binders, (e.g., PVC or chitosan) therefore overcoming the adhesion issues. [91][92][93][94][95][96] Alternatively, an adhesive thin film can be applied prior to the drop casting procedure encapsulating the particles (e.g., polypyrrole [97] and agarose [98,99] ), though the encapsulation may decrease the sensitivity of the receptor.
Recently, a novel optical-chemical sensor has been developed by Cennamo et al. demonstrating the flexibility of drop-casting. In this work, a 1 mm plastic optical fiber was used as a solid substrate; two different configurations with one or three microholes were obtained by drilling into the exposed core of the fiber. Afterward, the MIP prepolymerization mixture was drop cast in the holes and a thermal polymerization was carried out to obtain MIP microparticles (Figure 4). The obtained platform showed remarkable selectivity and sensitivity by SPR analysis. [100] As the method facilitates the incorporation of presynthesized receptor particles, drop-costing can be applied across a large range of potential targets and readout technologies ( Table 2).

Dip Coating
The dip coating technique is a simple and low-cost method to obtain thin-film coatings on a substrate. The process can   be divided into three main stages: dipping, withdrawal, and evaporation ( Figure 5). Since MIP particles synthesized by bulk polymerization are usually in the micrometer size range and thus tend to precipitate when suspended, their deposition via the classical technique is difficult to achieve. Therefore, to achieve the deposition of MIP microparticles via dip coating the particles need to be mixed with another adhesive component; in previous work, this was achieved by combining the microparticles with an agarose solution, followed by crosslinking treatment. [103] In recent years, the technique has been employed to functionalize thermocouples with nanoMIPs as synthetic receptors. The functionalized substrates were used as thermal sensors for the detection of various targets, such as small molecules [104] and proteins. [88,105] www.advancedsciencenews.com www.advsensorres.com

Direct Electrode Incorporation/Screen Printing
As electrochemical readout technologies become ever more popular, so has the appeal of integrating MIPs directly into electrodes that compliment these platforms. Multiple approaches have been attempted ranging from electropolymerization to in-situ polymerization, though more simplistic mechanical approaches have shown great promise. [36] One of these approaches utilizes bulk polymerized MIP particles and incorporates them directly onto an electrode's surface by use of carbon paste. [106] The MIP particles are directly mixed with the paste, enabling the distribution of the receptors across the entire electrode. Applications of this approach facilitate the detection of a wide range of potential analytes including, but not limited to, small organic molecules [107][108][109][110][111][112][113][114][115][116][117][118][119][120][121][122] and ions. [123][124][125] Supplementary materials can be added to the paste, boosting the conductivity of the composite and enhancing the sensitive of the system, with common additives including carbon nanotubes [108,110,111,124,125] or platinum nanoparticles. [113] In recent years, this concept has been taken even further with advancements in the field of screen printing. Rather than applying the MIP loaded graphene paste to a pre-existing electrode, instead the MIP particles are combined with graphene ink and screen printed directly as electrodes. [121,122] Thus, material waste is greatly reduced, electrode homogeneity is increased, and the substrate that the electrode is printed onto can be engineered towards more environmental friendly materials. [126] Furthermore, this has opened up the possibility of developing facile flexible electrochemical sensors that can be tailored toward a specific application in terms of substrate material, receptor, and print design.
The cost of this freedom comes in the form of diminished binding capacities and reduced linear ranges when exposed to a desired target analyte. As the MIP particles are combined with either a paste or ink, the surface of the receptor becomes saturated with graphene preventing a target analyte interacting with the specific binding sites. Thus, the essential blocking of the receptor reduces the aforementioned linear range and selectivity of the sensor, though the sensitivity remains somewhat constant as graphene is excellent at absorbing species.

Electrochemical MIP Deposition
Electrochemical deposition methods are categorized by their use of electric currents as either a way of applying force to repel/attract ions or to deposit ionic material onto an electrode. Thus, electropolymerization and electrospinning are the two main methods that embody this principle for the deposition of MIPs. As such, mass is added across a substrate in the form of fibers that contain MIPs or by thin film formation respectively. [127]

Electropolymerization
Electropolymerization is a deposition method in which a conductive polymeric coating/film is formed across a conductive substrate/electrode when an electroactive monomer is subjected to potentials that cause its oxidation and reduction. [25] More commonly, this approach is associated with the direct formation of imprinted layers at a substrate's surface, allowing imprinting during the deposition process by simply introducing a template species to the polymerization mixture. [128] This straightforward alteration has spawned an entire sector in the field of MIPs, proving popular for the imprinting of small molecules and larger (bacteria and proteins) species. [129,130] This process is easily achieved using voltammetric, [131] galvanostatic, [132] and potentiostatic conditions, [133] though voltammetric electropolymerization has proven to be the far more popular approach. In essence, a monomer is oxidized and reduced by sweeping across a set range of potentials, via cyclic voltammetry, stimulating polymerization. The thickness of the resulting polymeric layer is tuned by altering the sweep/scan rate, thus affecting the rate at which the polymer forms on a substrate. A three-electrode setup is the standard for this deposition approach, consisting of a working (where the coating is deposited), counter, and reference electrode (typically Ag/AgCl or saturated calomel electrode) (Figure 6). [134] The range of the potential sweep is monomer-dependent, with the redox potential differing depending on the compound selected. [136] The most common compounds to undergo electropolymerization are pyrrole, [137] analine, [138] and dopamine, [139] with these compounds being amongst the easiest to polymerize. Schwieger et al. highlight this by successfully preparing polypyrrole layers imprinted with clofibric acid by means of cyclic voltammetry. [140] The layers are deposited onto gold coated wafers (working electrode) by cycling the applied potential between -0.2 and 0.8 V across an aqueous solution containing clofibric acid, KNO 3 and PBS. After the synthesis, the layers were washed with 70% ethanol and a solution of potassium chloride/hydrochloric acid to remove unreacted monomer and the template compound. The resulting layer's physical characteristics were then studied, determining the response towards clofibric acid, the selectivity, the hydrophobicity by contact angle, and the layer thickness. AFM imaging revealed that the deposited material had a circular structure <1 m (diameter) on the surface, with a surface roughness of between 6 and 8 nm. Overall, this piece of research highlights the facile nature of the approach, with its applications being far spread and applied to many different molecules across the field ( Table 3).
The main downside to electropolymerization lies in the fact that electroactive monomers have to be utilized for layer formation/deposition. This means that currently there are a limited number of viable options that have been used to demonstrate MIPs can be deposited in this fashion. The limited library proves problematic when selecting a monomer that has optimum interactions with the template species, meaning it is hard to specifically tailor an electropolymerized layer towards a desired analyte. This said, custom monomers are slowly emerging that have greater structural diversity and offer more in terms of tailored specific interaction and stem from the field of traditional conjugated polymers. [154] Another downside is the substrate that the material is being electropolymerized onto has to be conductive, and even this is a challenge as metal such as copper and aluminum can be easily oxidized during the electrodeposition meaning that commonly gold, platinum, and carbon are the materials of choice. Figure 6. A) An electropolymerization setup, showing the critical components necessary, followed by B) a typical cyclic voltammograms (CV) that is collected during the electropolymerization process. Reproduced with permission. [135] Copyright 2022, Royal Society of Chemistry (CC-BY).

Electrospinning
Electrospinning is a method of producing fibers (micronanoscale) by means of applying an electric force to a charged polymer thread that is generated in a melt or solution and drawing them out into the desired dimensions. [155] In essence, a polymer solution is pumped through a charged needle. As the solution passes through, the liquid's surface becomes charged and the electrostatic repulsion begins to overcome the surface tension of the solution. As the solution leaves the needle, it is ejected towards a collection plate/drum that rotates and draws the solution near. As the solution flies toward the collector, the solvent evaporates and the remaining polymer is drawn out forming fibers (Figure 7). [156] Depending on the physical characteristics of the polymer solution, the electric field applied, drum rotation, and injection speed, it is possible to tailor the process towards a desired fiber size and morphology. [158] As the method is solution based, prior to spinning, MIPs can be dispersed into the liquid phase, yielding electro-spun fibers with MIPs distributed throughout the produced fibers. Mild spinning conditions also favor the use of this approach, as the MIP particles are not exposed to high temperatures or other damaging conditions. [159] Demirkurt et al. demonstrate this principle by encapsulating benzyl paraben imprinted polymers in styrene, before studying the extraction capabilities of the prepared fibers by analyzing spiked sea, tap, and bottled water samples. [160] The prepared MIP particles were implemented into the polymer fibers by mixing MIP particles into a polystyrene (10% w/v) solution in DMF, before electrospinning fibers using a 30 KV applied voltage across a 15 cm distance at 1.0 mL h −1 , yielding nanofibers. This methodology has been widely adopted by the MIP community with many variations of this process possible, utilizing different applied voltages, plate distances, flow rates, and matrix polymer compositions. An overview of literature referring to these modifications can be found in Table 4, outlining what parameters have been changed to achieve the incorporation/deposition of the MIP particles inside the fibers generated.
Alternatively, the molecular imprinting process can occur during the electrospinning and subsequent deposition of the material. [175] In contrast to the previous iteration, this approach enables the entirety of the electro-spun surface to act as a receptor rather than relying on particles distributed throughout the bulk of the material. Introducing a template species into the prespun polymeric solution allows for the formation of imprints within the fibers after spinning. [176] The downside of "in-situ spinning" being that the deposited layer is not highly crosslinked, which is normally definitive of MIPs as this instills rigidity and receptor stability. This said, strong ionic interactions stabilized by a template species have the potential to mimic this highly  [146] Further examples: [147][148][149][150][151][152][153] crosslinked characteristic and fibers spun have mimetic capabilities, or post crosslinking of the polymer are a possibility. [177] Huang et al. highlight this approach as they form nanofiber films for the sensitive detection of 2,4,6-tribromophenol (TDP) by spin coating a solution containing TDP alongside the monomercyclodextrin ( -CD) and poly-vinylbutyral (PDB) as electrospinning matrix. [178] The mixture was placed in a 10 mL syringe fitted with a metallic needle of 0.4 mm inner diameter, and was fitted horizontally opposing a stainless steel electrode that was directly connected to a high-voltage power supply. Between the tip and collector was applied 18 kV, with the distance between the two being 10 cm, and the flow rate of the solution 0.5 mL h −1 . The TDP doped polymer fibers were directly electrospun onto a polished glassy carbon electrode (GCE), and postcuring of the polymeric fibers was achieved by immersing in hexamethylene diisocyanate (HMDI) at room temperature for 24 h (Figure 8). The resulting layer showed great sensitivity towards the TDP, having a linear range between 0.9-10 × 10 -9 m and a calculated LoD of 0.629 × 10 -9 m when analyzed with a quartz crystal microbalance (QCM). Further examples can be found in Table 4, highlighting the versatility of the approach.
The primary downside to this technique is that the imbued MIP particles are not homogenously spread throughout the spun fiber, leading to sensor reliability issues. Furthermore, depending on the size of the particles that are captured in the fiber, surface area can be lost due to high levels of encapsulation and particles being buried. Thus leads to the further limitation of there being a maximum amount of MIP particles that can be mixed with the spun polymer matrix, with excessive amount resulting in poor fiber production and reduced mechanical properties.

Chemical Deposition
One of the major challenges associated with the development of sensing platforms is the reproducibility of the method employed to fabricate such sensors. In many MIP-based sensing technologies, the preferred approach to prepare the imprinted polymer is bulk polymerization, due to its ease and scalability of polymer Figure 7. Electrospinning schematic of imprinted nylon 6 nanofibers for the extraction of bisphenol A from waste water. Reproduced with permission. [157] Copyright 2018, John Wiley and sons. Further examples: [172][173][174] synthesis. Although bulk polymerization represents a rapid and simple preparation method, the integration of bulk MIP particles into sensing devices could lead to sensors with poor repeatability and high batch-to-batch variance. [179] In addition, it has been shown by different sources that MIP films lead to better performances when compared with bulk MIPs-based platforms that hold the same chemical composition. [97,180] The ways in which this can be utilized for MIP deposition are through the generation of thin films or the direct grafting of a receptor.

Thin Films
Thin films, also described as "in-situ polymerization" or selfassembly polymerization, [181] are achieved by adding (usually by drop casting technique) the prepolymerization mixture on the surface of a substrate and subsequently polymerized (Figure 9). The focus is the formation of MIP films across different substrates via UV light or thermal initiation, leading to the direct formation/deposition of a receptor. In recent years, the development of MIP thin films has gained increasing attention, stemming primarily from their easy incorporation into many readout technologies. [182] Since the polymer film can be designed with several monomers and/or cross-linker molecules, the variation of one of mixture components can have a huge impact not only on the sensitivity of the sensor but also on the transducer used to convert a given signal. For this reason, these films have found application with several readout technologies such as: HTM, [76,97,183] EIS, [76] CV, [184,185] DPV, [186][187][188][189] QCM, [70,[190][191][192] GC-MS, [193,194] SPR, [195,196] APGC-MS/MS, [197] UHPLC-PDA, [198] UV-Vis. [199] www.advancedsciencenews.com www.advsensorres.com Figure 8. SEM morphology of nanofibers a) before crosslinked; and b) after crosslinking. Reproduced with permissions. [178] Copyright 2016, MDPI (CC-BY).

Figure 9.
Fabrication scheme of MIP films for malathion detection on screen printed gold electrode. Reproduced with permissions. [188] Copyright 2020, Elsevier.

Surface Grafting
Similarly to thin films, surfacing grafting allows MIPs to be generated at a functionalized interface in-situ. This concept is similar to that of electropolymerization, though is defined by the deposited/formed layer being covalently bound to the substrates surface whereas electropolymerized layers rely on adsorption instead. [200] This is made possible by the prior functionalization of the substrate, where reactive chemical functionalities (e.g., hydroxyl, thiol, amine etc.) are introduced across the surface by simple chemical modification. [201] Linker molecules are proceeded to be coupled to the functionalized surface, offering "anchor" points for the growing of the polymeric layer. Typically, these anchors tend to be monomers or photo/thermal initiators that are easily incorporated into the deposited layer during the grafting process, thus offering a covalent linkage directly into the bulk of the deposited material. [202] Sulitzky et al. highlight this approach in their work, where they covalently bound 4,4'-azobis(4-cyanopentanoic acid) (ACPA) to (3-aminopropyl)triethoxysilane (APS) functionalized silica particles. [203] Subsequently, a methacrylic acid (MAA) and ethyleneglycol dimethacrylate (EGDMA) based MIP for the detection of D-phenylalanine/anilide was grafted by exposing the functionalized silica particles to these components in toluene while under constant irradiation from a UV light source. Alternatively, Tian et al. highlight how tetraethyl orthosilicate (TEOS) can be bound to hydroxylated steel sheets (10 × 20 mm) enabling the grafting of a MMA/EGDMA based molecularly imprinted polymer for the detection of enrofloxacin. [204] These approaches have been adopted by many researchers in the field, with examples listed in Table 5.
Due to the relative ease of synthesis, control over layer growth, and general tunability of the deposition method, the direct www.advancedsciencenews.com www.advsensorres.com Uric acid electrochemical Substrate: graphite electrodeSurface modification: sol-gel [207] Theophylline electrochemical Substrate : carbon nanotubesSurface modification: HydroxylationLinker: 3-chloropropyl trimethoxysilane [208] L-phenylalanine Electrochemical Substrate: silicaSurface modification: hydroxylationLinker : APS 4,4'-azobis(4-cyano pentanoic acid) bound [209] Glucose Potentiometric Substrate: Gold nanoparticles deposited on a screen printed carbon electrodeSurface modification: Benzoic acid functionalized poly(terthiophene)Linker: amide bond formation [210] Further examples: [211][212][213][214] surface grafting of MIPs is a highly attractive proposition. [215] The method also leans into scalability, with the potential of functionalizing a large surface and cutting them into smaller sensor size pieces with this notion theoretically generating sensors with higher homogeneity. Adversely, this functionalization has to be conducted in a controlled reproducible manner that could prove labor intensive and hard to replicate.

Vacuum Deposition Methods
Vacuum deposition is characterized as a group of processes that deposit materials species-by-species (e.g., atom-by-atom or molecule-by-molecule) onto a solid surface (or substrate) by operating at pressures below that of the atmosphere (e.g., in vacuum). Material that requires depositing is vaporized and introduced to a substrate, enabling direct layer formation across the surface. As these processes occur in the absence of gaseous species that would normally interfere with molecular or atomic deposition, it is possible to form highly controlled thin films onto the substrate. This class of deposition can be broken down into multiple subclasses including thermal evaporation, sputtering, cathodic arc vaporization, laser ablation, and chemical vapor deposition (CVD). [216][217][218][219] Though these different branches of vacuum deposition exist, only CVD has been reported for use in the deposition of MIPs.

Chemical Vapor Deposition
Traditionally CVD is a deposition method by which a vacuum draws volatile precursor compounds across a substrate (typically a wafer), where they can then react and deposit at the surface. [220] The resulting layers are highly robust and most importantly tenable, leading to this process being adopted mainly by the microfabrication industry for the production of semi-conductors. [221] Common materials that are currently deposited this way include silicon, carbon, metals (e.g., titanium, tungsten), and fluorocarbons, though polymer composites are becoming increasingly popular. [222] Different CVD operating conditions are selected Figure 10. The graphical concept of using iCVD for the direct synthesis of MIPs onto functionalized substrates. Reproduced with permissions. [219] Copyright 2013, American Chemical Society.
based on the material being deposited, with variations regarding the working pressure, vapor characteristics, and substrate heating being just a few parameters that are altered when making this consideration. [223] Currently the most common variation associated with MIPs is initiated chemical vapor deposition (iCVD), which is a method that is traditionally related to the deposition of dense conformal coatings onto solid substrates. [224] In essence, this method uses free-radical polymerization to synthesize thin films across a substrate, where the vaporized constituent components of the polymer (monomer, crosslinker, and initiator) are introduced inside a vacuum chamber containing template-functionalized substrates. The polymerization process is activated thermally by means of a heated filament, radicalizing the differing components and enabling them to react at the surface of the substrate. A cooling stage helps facilitate this step, being placed directly under the functionalized substrate, with an exhaust in close proximity to remove any unreacted material. [225] Ince et al. exemplified this approach by functionalizing anodic aluminum oxide membranes with immunoglobulin (IgG) before using iCVD for the fabrication of imprinted layers (Figure 10). [226] This research demonstrates how a complex biological target can be imprinted effectively by utilizing iCVD, producing highly selective receptors when exposed to other biological markers such as lysozyme and bovine serum albumin. Though highly effective, this approach has some major drawbacks, including the amount of parameters that must be optimized to achieve the synthetic receptor deposition alongside the associated high cost. CVD re-www.advancedsciencenews.com www.advsensorres.com quires highly specialized equipment, meaning the initial startup price for production is high and therefore making it less attractive than other more cost-effective approaches. On the other hand, this approach could prove extremely scalable and has the potential of producing very reproducible layers. The method however requires more study in terms of MIP-based sensor fabrication as this is the only study to date.

Conclusions and Future Outlook
As with the synthesis of imprinted polymers, there are many deposition approaches available for the integration of the resulting synthetic receptors into sensory platforms. The methods highlighted above, are currently the most widely accepted, with the versatility and the limiting factors of each approach discussed.
This said, there is still a vast amount of differing approaches that have not yet been investigated that could prove of high value. A large number of these methods remain in the field of vacuum deposition with thermal evaporation, sputtering, cathodic arc vaporization, and laser ablation remaining untouched and even CVD sparingly utilized in comparison to other approaches. The defining factor for deposition revolves around the idea that either an imprinted layer is formed in situ during sensor fabrication, or if the receptor has been synthesized prior and is being deposited as part of a bulk material. This is reflected in the target analytes each method is associated with, with in situ methods being more favorable for macromolecular targets such as bacteria, cells, and proteins. Whereas receptor particle deposition is more frequently associated with small molecules.
In terms of future prospects for the receptor deposition field, the most valuable advancements will be the ones that enable the scaling, increased reproducibility, and eventual commercialization of synthetic receptors. Of the approaches mentioned currently, electrodeposition and surface grafting offer the highest theoretical scalability, though each has its associated challenges that must be overcome. To date, there are very limited studies conducted on scaling these technologies with the vast majority of current research aimed at developing sensor toward novel analytes rather than focusing on this issue. This is not to say that researchers are not investigating potential applications of their sensors, but instead use applications to justify analyte selection rather than progressing the field towards more consistent and dependent sensor construction. It can therefore be imagined that the deposition of receptors will become a more crucial factor as time presses toward, with this being a vital stepping-stone for the advancement of the field and sensor technology as a whole.