Programming selectivity in molecularly imprinted polymers

Introduction

Selective capture and detection in chemically crowded media are persistent problems in analytical chemistry, separation science, remediation, food safety, biomedicine, and process engineering. Conventional sorbents and extraction platforms often fail when target species are present at trace levels, when structurally similar interferents coexist, or when matrix components such as salts, proteins, lipids, pigments, organic matter, and co-contaminants compete for binding. Molecularly imprinted polymers (MIPs) and ion-imprinted polymers (IIPs) address this challenge by forming recognition environments around a molecular or ionic template and preserving complementary cavities after template removal. The strongest contemporary evidence, however, suggests that imprinting should no longer be viewed only as a cavity-forming strategy; rather, it should be understood as a programmable soft-material approach in which recognition chemistry, matrix compatibility, architecture, operation, and readout are intentionally aligned.

The breadth of target classes illustrates this evolution. IIPs have been repeatedly developed for hazardous metals and metalloids such as lead (Pb²⁺), cadmium (Cd²⁺), arsenic (As³⁺), chromium species, thallium species, mercury species, and selenium species in water, wastewater, airborne particulate matter, food, biological materials, and industrial aqueous matrices (Abdullah et al., 2019; Abdullah et al., 2020; Abdullah et al., 2023; Aboufazeli et al., 2014; Amran et al., 2024; Arbab-Zavar et al., 2011; Buhani et al., 2010; Lesniewska et al., 2016; Li et al., 2011; Liu et al., 2013; Mafu et al., 2016; Rahman et al., 2017; Zhang et al., 2014). The same logic extends to recoverable, process-critical, or separation-intensive species, including silver [Ag(I)], lanthanum (La³⁺), ruthenium [Ru(III)], copper [Cu(II)], phosphorus species, and radioyttrium, where the objective shifts from contaminant control toward selective separation, purification, resource recovery, or radiochemical separation (Jalilian and Taheri, 2018; Ritt et al., 2019; Sarabadani et al., 2013; Taghizadeh and Shafizadeh, 2025; Wang et al., 2016; Zambrzycka et al., 2013; Zhang et al., 2017). Organic-template imprinting has similarly supported selective extraction or monitoring of pesticides, herbicides, endocrine-disrupting phenols, pharmaceutical compounds, mycotoxins, polybrominated diphenyl ether flame retardants, carbamate pesticides, and fuel-sulfur contaminants, while related selective adsorbent studies address per- and polyfluoroalkyl substances, dyes, and volatile H₂S across environmental, food, biological, and industrial systems (Akbarzadeh et al., 2022; Andrade et al., 2023; Borui et al., 2024; Danmaraya et al., 2025; Geng et al., 2024; Herrero-Hernández et al., 2013; Liu et al., 2022; Mahabadi et al., 2025; Marc et al., 2018; Mokhtari et al., 2022; Sun et al., 2020; Wang et al., 2021; Yang et al., 2011).

At the design level, programmability is introduced through several interacting levers. Ion-imprinted systems encode metal–ligand coordination environments and ligand identity during polymerization, as shown for Pb(II), Tl(III), Cr(III), Cu(II), Hg(II), Ag(I), Ru(III), and related ion-imprinted networks (Ambarsari et al., 2023; Arbab-Zavar et al., 2011; Fazelirad et al., 2014; Jalilian and Taheri, 2018; Lesniewska et al., 2016; Rahman et al., 2017; Taghizadeh and Shafizadeh, 2025; Zambrzycka et al., 2013; Zhang et al., 2014). Organic MIPs rely on template–monomer preorganization, hydrogen bonding, π–π interactions, electrostatic interactions, and complementary functional groups, often combined with crosslinked networks or bulk, precipitation, electropolymerization, reversible addition–fragmentation chain-transfer polymerization, and emulsion-templated routes (Akbarzadeh et al., 2022; Ameli and Alizadeh, 2012; Andrade et al., 2023; Ceglowski et al., 2017; Guc et al., 2021; Ji et al., 2013; Mantashloo and Bahar, 2025; Piryaei et al., 2021; Savigni et al., 2025; Silva et al., 2018; Soledad-Rodríguez et al., 2017). Increasingly, chemical recognition is integrated with physical architecture: magnetic cores enable rapid retrieval from complex samples; silica, graphene/carbon, and mesoporous supports improve binding-site accessibility; and hybrid systems combine imprinting with aptamers, boronate affinity, zeolitic imidazolate framework- and metal–organic framework-type materials, photocatalysts, or nanostructured transducers (Ciarlantini et al., 2025; Dinali et al., 2021; Gao et al., 2024; Geng et al., 2024; Piryaei et al., 2021; Shahar et al., 2018; Sun et al., 2020; Tonucci et al., 2020; Wei et al., 2018; Xu et al., 2013).

The performance literature confirms that imprinted soft materials are most valuable when selective recognition is converted into measurable function. Representative studies report low detection limits for metal ions and organic contaminants, high adsorption capacities for selected inorganic and organic targets, strong selectivity over competing ions or structural analogues, and enrichment gains that improve analytical sensitivity (Abdullah et al., 2019; Abdullah et al., 2020; Abdullah et al., 2023; Akbarzadeh et al., 2022; Buhani et al., 2010; Khoddami and Shemirani, 2016; Kumar et al., 2020; Mahabadi et al., 2025; Mokhtari et al., 2022; Roushani et al., 2018; Zambrzycka et al., 2013; Zhang et al., 2014). Reusability and operational durability further distinguish practical systems from proof-of-concept sorbents, with selected microextraction by packed sorbent, magnetic extraction, cryogel/composite, monolith, and thermoresponsive platforms demonstrating repeated use over multiple cycles or extended extraction sequences (Andrade et al., 2023; Bhogal et al., 2021b; Dil et al., 2021; Dong et al., 2014; Khoddami and Shemirani, 2016; Pantanit et al., 2025; Rahman et al., 2017; Sartore et al., 2020; Venkatesh et al., 2014; Wang et al., 2016; Yang et al., 2012). Device-compatible formats, including imprinted electrochemical and electrical sensors for paracetamol and moxifloxacin, extend the concept toward transduction-ready recognition layers capable of reversible and selective signal generation under interferent challenge (Shafique et al., 2024; Memoona et al., 2024).

Real-sample validation is therefore central to the field’s translational claim. Environmental waters, food and beverage matrices, biological fluids, industrial streams, gas mixtures, fuel samples, and radiochemical targets collectively show that imprinted soft materials can operate beyond idealized laboratory solutions (Abril et al., 2026; Andrade et al., 2023; Chen et al., 2022; da Silva et al., 2017; Herrero-Hernández et al., 2013; Mahabadi et al., 2025; Orachorn and Bunkoed, 2022; Sarabadani et al., 2013; Sun et al., 2020; Wang et al., 2021; Yang et al., 2012; Zhang et al., 2017; Zhao et al., 2011). Yet this evidence also calls for disciplined interpretation. Several imprinted systems outperform non-imprinted polymers, activated carbon, polymeric adsorbent resin, octadecylsilane sorbents, Oasis hydrophilic–lipophilic-balanced sorbents, and liquid–liquid extraction in specific applications (Andrade et al., 2023; Buhani et al., 2010; Herrero-Hernández et al., 2013; Krupadam et al., 2012; Li et al., 2011; Önnby et al., 2012; Tonucci et al., 2020; Venkatesh et al., 2014; Zhao et al., 2011), but imprinting is not universally superior. Cases in which non-imprinted porous materials or low-cost electrochemical platforms perform competitively underscore the need to define where imprinting provides unique value rather than incremental complexity (Meng et al., 2022; Mushtaq et al., 2026). This review therefore positions programmable imprinted soft materials as a versatile but deployment-dependent class of recognition platforms whose significance rests on the alignment of target, matrix, function, and validation.

Target analyte and study objective

The studies reviewed in this section define the target scope and functional objectives of programmable imprinted soft-material platforms and related imprinted sorbent systems. Here, programmability refers to the deliberate matching of three elements: a defined target analyte or analyte class, a complex medium in which selective operation is required, and a preassigned function such as removal, adsorption, extraction, enrichment, preconcentration, separation, cleanup, purification, recovery, monitoring, detection, or determination. This framing distinguishes target-directed imprinting from nonspecific sorption and positions selectivity as a practical function that must operate in matrices containing competing ions, structurally related molecules, coextractives, salts, proteins, pigments, or process contaminants. Quantitative performance metrics are not discussed in this section because the purpose here is to establish analyte scope and study objectives rather than evaluate outcomes.

Inorganic ions, nutrient species, and radionuclides

IIP systems provide a direct illustration of target-programmed selectivity because their objectives are commonly defined by the need to isolate a specific ion from competitive ionic backgrounds. Toxic metal and metalloid species were prominent targets, including Pb²⁺, Cd(II)/Cd²⁺, arsenic species, Cr(III)/Cr(VI), Tl(I)/Tl(III), mercury species, and selenium ions in water, wastewater, river water, airborne particulate matter, food samples, human urine, industrial aqueous samples, and other environmental or biological matrices (Abdullah et al., 2019; Abdullah et al., 2020; Abdullah et al., 2023; Aboufazeli et al., 2014; Adauto et al., 2020; Amran et al., 2024; Arbab-Zavar et al., 2011; Buhani et al., 2010; Darroudi et al., 2020; Kumar et al., 2020; Lesniewska et al., 2016; Li et al., 2011; Liu et al., 2013; Mafu et al., 2016; Önnby et al., 2012; Panjali et al., 2015; Panjali et al., 2016; Rahman et al., 2017; Zhang et al., 2014). Across these studies, the central objective was selective removal, adsorption, preconcentration, speciation, separation, or determination of hazardous inorganic species in media where coexisting ions and matrix components can limit nonspecific capture.

The same target-directed logic was extended to valuable, recoverable, or process-critical inorganic species. Cu(II), Co(II), Fe(III), Zn(II), Ag(I), gold, Ru(III), La³⁺, and palladium-related imprinted systems were investigated for selective extraction, determination, recovery, or purification in environmental samples, aqueous media, food and biological matrices, radiology film, concentrated zinc sulfate solutions, hydrometallurgical streams, and other industrially relevant settings (Ebrahimzadeh et al., 2013; Fazelirad et al., 2014; Jalilian and Taheri, 2018; Khoddami and Shemirani, 2016; Lai et al., 2019; Özkahraman et al., 2011; Pardoyo et al., 2024; Taghizadeh and Shafizadeh, 2025; Wang et al., 2016; Zambrzycka et al., 2013; Zhang et al., 2017). Beyond conventional ion analysis, low-concentration phosphorus recovery from eutrophic surface waters and radioyttrium isotope separation from strontium targets broadened the objective from contaminant control to resource recovery and radiochemical separation (Ritt et al., 2019; Sarabadani et al., 2013). Together, these studies show that ion imprinting is used not only to remove hazardous species but also to recover useful targets and enable selective purification in chemically demanding systems.

Organic pollutants and industrial contaminants

A second major application domain involves organic pollutants and industrial contaminants. In these studies, imprinting is programmed around molecular identity, structural similarity, and matrix complexity. Pesticides and herbicides were frequently targeted, including bentazone, sulfonylureas, organophosphorus pesticides, fenthion, fipronil, parathion, carbamates, triazines, methoxyfenozide, diazinon, dinotefuran, phenylurea herbicides, and phenoxyacid herbicides in water, vegetables, corn, fruit juice, milk, apple juice, and broader agricultural or environmental analytical contexts (Akbarzadeh et al., 2022; Andrade et al., 2023; Danmaraya et al., 2025; Dinali et al., 2021; Geng et al., 2025; Hayat et al., 2022; Herrero-Hernández et al., 2013; Li et al., 2018; Miao et al., 2015; Pantanit et al., 2025; Phirisi et al., 2024; Piryaei et al., 2021; Rahiminezhad et al., 2010; Sa-nguanprang et al., 2024; Silva et al., 2018; Tang and Xiang, 2010; Wang et al., 2021). The shared objective across these studies was selective extraction, enrichment, cleanup, residue monitoring, or determination of agrochemical targets in media containing structurally related compounds and matrix-derived interferences.

Persistent organic, industrial, and process contaminants further demonstrate the translational range of imprinted or selectively adsorptive function. Targets included short-chain per- and polyfluoroalkyl substances, tetrabromobisphenol A, polybrominated diphenyl ethers, polychlorinated biphenyls, phthalates and phthalate monoesters, 4-nitrophenol, dichlorophen, Acid Green 16, methylene blue, Direct Red 81, pyridine, carbamazepine, dibenzothiophene, and hydrogen sulfide in wastewater, textile effluents, biological fluids, water, fuel oils, gas streams, complex environmental systems, and industrially relevant fuel- or gas-treatment media (Borui et al., 2024; Ciarlantini et al., 2025; Diab et al., 2021; Guo et al., 2019; Lang et al., 2022; Li et al., 2021; Li et al., 2024; Mahabadi et al., 2025; Marc et al., 2018; Mathidala and Ogunlaja, 2019; Moradirad et al., 2024; Mortari et al., 2021; Savigni et al., 2025; Venkatesh et al., 2014; Wang et al., 2025; Wei et al., 2018; Wolska et al., 2020; Yang et al., 2012; Yang et al., 2011). In this group, the objective extends beyond analytical extraction to broader selective functions, including environmental remediation, gas purification, fuel cleanup, direct solid-phase measurement, and removal of persistent or process-relevant contaminants.

Endocrine-disrupting and phenolic compounds represent a closely related subset of organic targets. Bisphenol A, bisphenol F, 4-cumylphenol, 17β-estradiol, phenols, and related phenolic contaminants were targeted for selective recognition, removal, extraction, or determination in foodstuffs, beverages, seawater, yogurt, milk, soft drinks, surface water, groundwater, and water-treatment contexts (Bhogal et al., 2021b; Dong et al., 2014; Herrero-Hernández et al., 2013; Jullakan et al., 2025; Mamman et al., 2021; Tonucci et al., 2020). These examples reinforce the importance of matching the imprinting objective to both analyte identity and matrix composition, particularly where hormonally active or phenolic compounds must be distinguished from chemically rich backgrounds.

Food, beverage, and agricultural safety targets

Food, beverage, and agricultural matrices provide a distinct test of selective function because target analytes often occur alongside endogenous coextractives and matrix-derived interferences. Imprinted materials were developed for residue determination, extraction, cleanup, enrichment, or contaminant removal in milk, honey, apple juice, fruit juice, wine, vegetables, corn, pork, soft drinks, and natural-product matrices. Representative targets included fipronil in milk, trichlorfon in vegetables, phenylurea and triazine herbicides in fruit or vegetable matrices, clenbuterol in pork, and multiresidue pesticides in apple juice or corn (Andrade et al., 2023; Dinali et al., 2021; Du et al., 2014; Hayat et al., 2022; Ji et al., 2013; Jullakan et al., 2025; Li et al., 2018; Pantanit et al., 2025; Phirisi et al., 2024; Sa-nguanprang et al., 2024). These studies position imprinting as a strategy for selective sample preparation and residue monitoring in food matrices where nonspecific extraction can reduce analytical reliability.

Food-related chemical safety also included antibiotics, pharmaceuticals, mycotoxins, beverage defects, and natural-product targets. Aminoglycoside antibiotics in honey, ampicillin in cow milk, wastewater-relevant fluoroquinolone antibiotics, and nonsteroidal anti-inflammatory drugs in milk were addressed through selective extraction, preconcentration, or determination objectives (Ji et al., 2013; Orachorn and Bunkoed, 2022; Prieto et al., 2011; Soledad-Rodríguez et al., 2017). Ochratoxin A, zearalenone, and patulin were investigated for selective enrichment, purification, determination, or removal in food and biological contexts, while m-cresol was targeted as a smoke-taint compound in red wine (Abril et al., 2026; Geng et al., 2024; Mausia et al., 2011; Sun et al., 2020; Zhao et al., 2011). Ellagic acid and quercetin were also targeted in agricultural waste or food matrices, extending the section’s scope from contaminant monitoring to selective recovery or detection of food-associated bioactive compounds (Gao et al., 2024; Mantashloo and Bahar, 2025).

Pharmaceutical, biomedical, clinical, and forensic targets

A further body of work applies imprinting to selective function in biological, pharmaceutical, clinical, forensic, and chemically competitive sensing media. Pharmaceutical and chiral targets included S-citalopram, carvedilol, amlodipine enantiomers, local anesthetics, sotalol, naproxen, clopidogrel, diclofenac, phenobarbital, theophylline, melatonin, paclitaxel, carbamazepine, and other therapeutic compounds in human urine, plasma, serum, physiological samples, milk, pharmaceutical samples, artificial gastrointestinal media, and related biological or environmental matrices (Alharbi et al., 2024; Ameli and Alizadeh, 2012; Ansari and Masoum, 2018; da Silva et al., 2017; da Silva et al., 2024; Daryanavard et al., 2013; Dil et al., 2021; Luo et al., 2011; Mokhtari et al., 2022; Orachorn and Bunkoed, 2022; Pebdani et al., 2015; Rahimi et al., 2022; Savigni et al., 2025; Tan et al., 2021). In these cases, the objective was selective extraction, preconcentration, enantioseparation, determination, or controlled delivery in systems where structural similarity, low abundance, and biological matrix effects complicate conventional analysis.

Clinical, forensic, diagnostic, exposure-monitoring, electrochemical sensing, and bioactive-molecule targets included testosterone, estrogens, parabens, cannabinoids and their metabolites, sarcosine, mandelic acid, phthalate metabolites, salicylic acid, 2′-deoxyadenosine, flavins, nicotine, linalool, propyphenazone, methylparaben, ellagic acid, quercetin, paracetamol, and moxifloxacin (Bui and Haupt, 2011; Ceglowski et al., 2017; de Oliveira et al., 2019; Gao et al., 2024; Li et al., 2021; Liu et al., 2022; Mantashloo and Bahar, 2025; Moein et al., 2015; Pajewska-Szmyt et al., 2020; Rahangdale et al., 2016; Rahangdale and Kumar, 2021; Sartore et al., 2020; Shahar et al., 2018; Soleimani et al., 2018; Wang et al., 2023; Shafique et al., 2024; Memoona et al., 2024). These studies shift the objective from contaminant removal toward selective information extraction, including biomarker determination, forensic testing, exposure monitoring, molecular recognition, biological-fluid analysis, gas-phase sensing, and electrochemical detection. As a group, they connect programmable imprinting to analytical, biomedical, and sensing settings in which the value of the material is defined by its ability to isolate a target signal from a complex or chemically competitive background.

Target–medium–function map

The studies summarized above are best understood through a target–medium–function framework, in which the study objective is defined by the relationship among the selected analyte, the complexity of the medium, and the intended selective function. This organizing logic is illustrated in Figure 1 and summarized in Table 1, which together clarify how the same imprinting principle is adapted across environmental, food, biological, industrial, radiochemical, and sensing settings.

TABLE 1. Target–medium–function patterns in programmable imprinted soft materials.

Imprinted material design and synthesis

In this section, “programmability” refers to the way selective function is encoded during material formation through template choice, monomer or ligand organization, crosslinking chemistry, support architecture, responsive components, and post-imprinting activation. The reviewed materials include conventional MIPs, IIPs, magnetic and core–shell composites, carbon- and silica-supported systems, chitosan- and cellulose-based matrices, conducting films, cryogels, monoliths, poly(high-internal-phase-emulsion) materials, hollow porous architectures, device-compatible sensors, and dual-recognition hybrids. Rigid or inorganic supports are included only where they organize, stabilize, or spatially present an imprinted soft recognition phase. Together, these systems show that selective function in complex media is programmed through five connected design levers: chemical recognition, spatial presentation of binding sites, matrix compatibility, external responsiveness, and activation of imprinted cavities. This synthesis-level programming logic is illustrated in Figure 2.

Table 2 summarizes these design levers without duplicating the narrative or introducing outcome metrics. Performance data such as adsorption capacity, selectivity coefficients, recovery, limits of detection, limits of quantification, reusability, and real-sample validation are intentionally not emphasized here because they belong to the outcomes and application sections. Instead, the table clarifies how synthesis choices create distinct forms of programmable behavior, including ion coordination, interfacial recognition, magnetic recovery, water compatibility, dummy-template substitution, stimulus response, morphology control, and device compatibility.

TABLE 2. Programmable design strategies in imprinted soft materials for selective function in complex media.

Chemical programming through template–monomer and ion–ligand organization

The most direct form of programmability is chemical: recognition sites are defined by the pre-polymerization interaction between the template and the surrounding monomers or ligands. In ion-imprinted systems, this was usually achieved by forming a coordination complex between the target ion and a ligand or functional monomer before polymerization. This strategy was applied to Pb(II), As(III), Cd(II), Tl(I)/Tl(III), Cr(III)/Cr(VI), Co(II), Ag(I), Cu(II), Zn(II), Hg(II), Ru(III), La(III), Fe(III), gold, arsenic, selenium, phosphorus species, and radioyttrium through bulk, precipitation, thermal, sol–gel, surface-imprinting, hydrothermal-assisted sol–gel, co-precipitation, grafting, and activators regenerated by electron transfer atom-transfer radical polymerization approaches (Abdullah et al., 2019; Abdullah et al., 2020; Arbab-Zavar et al., 2011; Darroudi et al., 2020; Ebrahimzadeh et al., 2013; Fazelirad et al., 2014; Jalilian and Taheri, 2018; Khoddami and Shemirani, 2016; Kumar et al., 2020; Lesniewska et al., 2016; Li et al., 2011; Liu et al., 2013; Mafu et al., 2016; Özkahraman et al., 2011; Pardoyo et al., 2024; Rahman et al., 2017; Ritt et al., 2019; Roushani et al., 2018; Sarabadani et al., 2013; Taghizadeh and Shafizadeh, 2025; Wang et al., 2016; Zambrzycka et al., 2013; Zhang et al., 2014). Representative examples include ternary-complex Pb(II)-IIPs prepared with 4-(2-pyridylazo)resorcinol, 4-vinylpyridine, methacrylic acid, and trimethylolpropane trimethacrylate, with nitric-acid leaching specified in one case (Ambarsari et al., 2023; Amran et al., 2024). Tl(I)/Tl(III)-IIPs were prepared from 5,7-dichloroquinoline-8-ol complexes using methyl methacrylate, ethylene glycol dimethacrylate (EGDMA), azobisisobutyronitrile (AIBN), and acetonitrile or 2-methoxyethanol as porogenic media (Arbab-Zavar et al., 2011; Darroudi et al., 2020). Cr(III)-IIPs were synthesized from Cr(III)–nicotinate complexes with acrylamide, EGDMA, and AIBN (Lesniewska et al., 2016), while Cu(II)-IIPs were prepared by bulk polymerization using a Cu(II)–1-(2-pyridylazo)-2-naphthol complex, methacrylic acid, and EGDMA, followed by template removal with 2 M HNO₃ (Fazelirad et al., 2014). Hg(II)-imprinted monoliths were obtained by thermally copolymerizing a mercury–cysteine-based ligand complex with methacrylic acid, EGDMA, and benzoyl peroxide in a polyethylene-straw mold (Rahman et al., 2017). These examples show that ion-imprinted selectivity is programmed through coordination environment, ligand identity, and crosslinked network formation rather than through the polymer backbone alone.

Organic-template imprinting followed the same general logic but relied mainly on template–monomer preassembly, hydrogen bonding, π–π interactions, electrostatic interactions, base-complementary pairing, or complementary functional groups, as reported in the summaries. Bulk, precipitation, suspension, surface, sol–gel, noncovalent, emulsion-templating, and reversible addition–fragmentation chain-transfer polymerization were used to prepare MIPs for pesticides, herbicides, pharmaceuticals, endocrine disruptors, mycotoxins, dyes, antibiotics, biomolecules, and bioactive or diagnostic analytes (Andrade et al., 2023; Guc et al., 2021; Hayat et al., 2022; Ji et al., 2013; Liu et al., 2022; Mantashloo and Bahar, 2025; Mausia et al., 2011; Piryaei et al., 2021; Prieto et al., 2011; Sartore et al., 2020; Savigni et al., 2025; Silva et al., 2018; Soledad-Rodríguez et al., 2017; Tan et al., 2021; Tang and Xiang, 2010; Wang et al., 2023). Reported monomers, crosslinkers, initiators, and porogenic components included methacrylic acid, methyl methacrylate, 4-vinylpyridine, acrylic acid, N-isopropylacrylamide, 2-vinylpyridine, eugenol, coumaric acid, itaconic acid, EGDMA, divinylbenzene, trimethylolpropane trimethacrylate, tetraethyl orthosilicate, AIBN, and 4,4′-azobis(4-cyanopentanoic acid) (Akbarzadeh et al., 2022; Ameli and Alizadeh, 2012; Ceglowski et al., 2017; Dong et al., 2014; Ji et al., 2013; Krupadam et al., 2012; Marc et al., 2018; Moradirad et al., 2024; Orachorn and Bunkoed, 2022; Rahiminezhad et al., 2010; Savigni et al., 2025; Silva et al., 2018; Wolska et al., 2020; Zhao et al., 2011). In several systems, computational or density functional theory-guided screening was used to guide monomer selection or template–monomer preassembly, including materials designed for carvedilol, salicylic acid recognition, palladium imprinting, carbamazepine adsorption, and dinotefuran recovery (da Silva et al., 2017; Lai et al., 2019; Rahangdale et al., 2016; Savigni et al., 2025; Silva et al., 2018). These examples indicate that chemical programmability is not limited to selecting a template; it also depends on how the template is paired with monomers, ligands, crosslinkers, and porogenic conditions during network formation.

Architectural programming through supports, interfaces, and hybrid structures

A second design lever is architectural: recognition sites are positioned on supports or within hybrid structures to improve accessibility, separability, or compatibility with complex media. Magnetic platforms were especially prominent because they combine an imprinted soft recognition phase with a recoverable magnetic core. Reported supports included Fe₃O₄, Fe₃O₄@SiO₂, Fe₃O₄@TiO₂@SiO₂, Fe₃O₄@SiO₂@CdTe, Fe₃O₄@SiO₂@chitosan–graphene oxide, magnetic graphene oxide, and cyclodextrin-modified magnetic mesoporous materials (Bhogal et al., 2021b; Ciarlantini et al., 2025; Dil et al., 2021; Khoddami and Shemirani, 2016; Kumar et al., 2020; Li et al., 2018; Pajewska-Szmyt et al., 2020; Sun et al., 2020; Taghizadeh and Shafizadeh, 2025; Wei et al., 2018). These architectures were used for tetrabromobisphenol A, bisphenol A, melatonin, trichlorfon, parabens, patulin, phthalates, carbamate pesticides, Cd(II), Co(II), Cr(VI), and Ag(I), among other targets (Ciarlantini et al., 2025; Dil et al., 2021; Li et al., 2018; Mamman et al., 2021; Pajewska-Szmyt et al., 2020; Sun et al., 2020; Taghizadeh and Shafizadeh, 2025; Wang et al., 2019; Wang et al., 2021; Wei et al., 2018). In these systems, programmability is not limited to recognition chemistry; physical separability is built into the material design.

Carbonaceous and nanoscale supports similarly repositioned imprinting from bulk polymer domains to interfacial recognition environments. Cd(II)- and gold-imprinted polymers were prepared on multiwalled carbon nanotubes, graphene oxide-supported MIPs were prepared for norfloxacin, MIP-coated carbon microspheres were developed for dibenzothiophene, and silver nanoparticle-core MIPs consisted of a silver nanoparticle core coated with a thin linalool-imprinted shell (Aboufazeli et al., 2014; Chen et al., 2022; Ebrahimzadeh et al., 2013; Shahar et al., 2018; Yang et al., 2011). Silica and mesoporous supports provided another route to spatially accessible recognition layers. Surface-imprinted polymers were prepared on silica gel for clenbuterol, silica nanoparticles were used as mesoporous core supports for magnetic MIPs targeting pesticide residues, SiO₂-COOH carriers were used for fenthion, SBA-15 was used for methoxyfenozide through reversible addition–fragmentation chain-transfer polymerization, and silica beads were used for patulin imprinting with 6-hydroxynicotinic acid as a template substitute (Dinali et al., 2021; Du et al., 2014; Geng et al., 2025; Piryaei et al., 2021; Zhao et al., 2011). These supported formats encode selectivity at interfaces where targets in complex fluids or gas-phase sensing systems can access recognition sites more directly than in dense bulk polymers.

Hybrid systems added further functional layers to imprinting. Zeolitic imidazolate framework–boronate-affinity MIPs combined boronate-affinity recognition with imprinting for ellagic acid, while zeolitic imidazolate framework-67/gold nanoparticle/aptamer/MIP materials combined aptamer immobilization and molecular imprinting for ochratoxin A (Gao et al., 2024; Geng et al., 2024). MIL-101(Cr)@MIPs were designed as core–shell hybrid nanosorbents for H₂S recognition, and MIP-TiO₂ materials were prepared using 17β-estradiol as the template (Mahabadi et al., 2025; Tonucci et al., 2020). Hybrid ionic imprinted polymers were also synthesized from 1-vinylimidazole, organosilanes, trimethylolpropane trimethacrylate, AIBN, and tetraethyl orthosilicate (Adauto et al., 2020). In these cases, the imprinted polymer is not a standalone sorbent; it is integrated with porous, catalytic, affinity-based, or inorganic components that add a second design function while preserving template-directed recognition.

Soft, biopolymer-based, and dummy-template matrices for complex media

Soft and water-compatible matrices are especially relevant to selective function in complex media because they embed recognition chemistry in bio-derived, hydrophilic, or flexible polymer environments. Chitosan-based MIP gels were crosslinked with glutaraldehyde using m-cresol as the template, cellulose-graft-tetraethylenepentamine MIPs were prepared using 4-nitrophenol, and chitosan or acrylamide-grafted chitosan systems used 4-hydroxybenzoic acid as a dummy or mimic template for salicylic acid recognition (Abril et al., 2026; Lang et al., 2022; Rahangdale et al., 2016; Rahangdale and Kumar, 2021). A phenobarbital-imprinted solid-phase microextraction fiber was prepared by coating chitosan onto stainless-steel wire with glutaraldehyde as the crosslinker and phenobarbital as the template (Rahimi et al., 2022). Aminoethyl chitosan was also used in a magnetic Cd(II)-IIP design by coating the polymer onto Fe₃O₄@SiO₂ nanoparticles and imprinting Cd(II) ions on the surface (Wang et al., 2019). These examples place imprinting within softer or more hydrophilic matrices while retaining target-directed site formation.

Dummy-template and mimic-template strategies provide another form of programmable design because the imprint is generated using a structural analogue rather than the final target itself. This approach was used where direct imprinting could be limited by template leakage, toxicity, cost, or unfavorable template behavior, or where a substitute template improved practical imprint formation. Reported examples include diethylphosphonoacetic acid for organophosphorus pesticides, diallyl phthalate for dibutyl phthalate, 4,4′-dihydroxydiphenyl ether for polybrominated diphenyl ethers, 10-deacetylbaccatin III and a paclitaxel side-chain template for paclitaxel, 4-hydroxybenzoic acid for salicylic acid recognition, and 6-hydroxynicotinic acid for patulin (Danmaraya et al., 2025; Li et al., 2024; Marc et al., 2018; Rahangdale et al., 2016; Tan et al., 2021; Zhao et al., 2011). This indirect imprinting strategy preserves the design aim of selective recognition while avoiding direct reliance on the target molecule as the imprinting template.

Responsive, porous, and device-compatible imprinted formats

Stimuli-responsive imprinted materials most explicitly express the programmable-materials concept because recognition, uptake, or release can be coupled to an external trigger. Thermosensitive bisphenol A-imprinted MIPs were prepared on porous polystyrene-derived carriers using 4-vinylpyridine and N-isopropylacrylamide, while thermally responsive mesoporous silica nanosphere MIPs were synthesized by two-step precipitation polymerization for dichlorophen (Dong et al., 2014; Guo et al., 2019). A pH-responsive diethyl phthalate-imprinted polymer was prepared from acrylic acid and methyl methacrylate crosslinked with EGDMA or divinylbenzene; one reported formulation contained 30% acrylic acid:methyl methacrylate at a 1:1 ratio, 70 wt% divinylbenzene, and 7 wt% template (Wolska et al., 2020). Photonic/magnetic dual-responsive Fe₃O₄@MIPs were constructed for caffeine by integrating stimuli-responsive polymers with molecular imprinting to enable light-regulated uptake and release (Xu et al., 2013). These systems connect imprinting with externally addressable behavior rather than passive binding alone.

Morphology-controlled materials programmed function through accessibility, permeability, or handling. MIP–poly(high-internal-phase-emulsion) foams were produced by emulsion templating for bentazone and clopidogrel (Akbarzadeh et al., 2022). Additional shape- or structure-directed systems included nicotine surface-imprinted monoliths, pyridine-imprinted poly(4-vinylaniline-co-divinylbenzene), porous poly(vinyl alcohol) cryogel composites, hollow porous MIPs, bowl-shaped hollow porous MIPs on polydopamine particles, and straw-supported La³⁺-IIPs prepared by surface imprinting with activators regenerated by electron transfer atom-transfer radical polymerization (Bhogal et al., 2021a; Huang et al., 2019; Mathidala and Ogunlaja, 2019; Pantanit et al., 2025; Wang et al., 2016; Wang et al., 2023). Conducting and device-compatible formats further extended imprinted materials beyond conventional sorbents. A naproxen-imprinted polypyrrole film was electrosynthesized by cyclic voltammetry and then overoxidized to generate recognition cavities (Ameli and Alizadeh, 2012). Silver nanoparticle-core MIPs were deposited on interdigitated array electrodes by drop casting or dip coating, combining an imprinted shell with a metal nanoparticle core and device-compatible transduction architecture (Shahar et al., 2018). Additional device-compatible MIP sensors were prepared by thermal free-radical bulk polymerization for paracetamol and moxifloxacin detection. The paracetamol-imprinted copolymer used methacrylic acid and styrene as functional monomers, EGDMA as the crosslinker, AIBN as the initiator, and dimethyl sulfoxide as the porogenic solvent, whereas the moxifloxacin-imprinted polymer used methacrylic acid, EGDMA, AIBN, and dimethyl sulfoxide and was integrated with interdigital electrodes as transducers (Shafique et al., 2024; Memoona et al., 2024). These examples show how synthesis can program not only what the material recognizes but also how the recognition sites are accessed, regenerated, or integrated into an operational format.

Template removal and activation of imprinted sites

Template removal converts preorganized template–monomer or ion–ligand assemblies into accessible recognition sites, but this procedure was not consistently described across the summaries. Where removal details were reported, acid-based steps were common. S-citalopram was removed from imprinted polymer particles by acidic extraction (Alharbi et al., 2024). Pb(II)-IIPs were leached with 5 M HNO₃, Tl(I)-IIPs with 1 mol/L HNO₃, and Cu(II)-IIPs with 2 M HNO₃ (Ambarsari et al., 2023; Darroudi et al., 2020; Fazelirad et al., 2014). In conducting MIPs, recognition sites were generated by overoxidation of the naproxen-imprinted polypyrrole film rather than by conventional solvent or acid leaching (Ameli and Alizadeh, 2012). For many magnetic, hybrid, cryogel, supported, and conventional MIP systems, the summaries identify the imprinting strategy and material architecture but do not specify the post-synthesis template-removal procedure; therefore, no removal chemistry is inferred. This distinction is important because the accessibility of imprinted sites depends on template removal or activation after synthesis, yet the available summaries report that step unevenly across material classes.

Characterization and experimental conditions

This section examines how programmable imprinted soft-material and hybrid platforms were verified, optimized, and operationally deployed for selective function in complex media. The discussion is restricted to characterization techniques, computational or theoretical tools, material architectures, optimized experimental variables, extraction or adsorption conditions, and analytical workflows. Analytical performance outcomes, including limits of detection, limits of quantification, linearity, recovery, relative standard deviation, adsorption capacity, removal efficiency, selectivity coefficients, and real-sample validation results, are excluded unless a value directly describes material verification or an experimental operating condition.

From imprinted architecture to verified programmable structure

The characterization evidence indicates that selective function was not established by imprinting chemistry alone, but by the deliberate construction of verifiable material architectures. Recognition sites were introduced into polymeric, silica-supported, magnetic, carbon-based, chitosan-derived, cellulose-based, cryogel, mesoporous, metal–organic framework-supported, zeolitic imidazolate framework-supported, and hybrid matrices. The verification strategy was therefore shaped by the intended operating environment. Scanning electron microscopy (SEM) and Fourier-transform infrared spectroscopy (FTIR) were the most consistently used methods for confirming morphology and functional-group incorporation, whereas energy-dispersive X-ray spectroscopy, X-ray diffraction (XRD), thermogravimetric analysis, Brunauer–Emmett–Teller (BET) analysis, elemental analysis, nuclear magnetic resonance spectroscopy, transmission electron microscopy, field-emission scanning electron microscopy, high-resolution transmission electron microscopy, thermogravimetric/differential thermal analysis, differential scanning calorimetry coupled with thermogravimetric analysis, vibrating-sample magnetometry, X-ray photoelectron spectroscopy, nitrogen adsorption–desorption analysis, and porosimetry were used when additional evidence of composition, crystallinity, surface area, porosity, thermal stability, magnetic behavior, or interfacial structure was required.

This verification framework directly supports the logic of programmable imprinted soft materials: the programmed recognition environment must be structurally and chemically demonstrable before it can be connected to selective operation in a complex medium. Metal-ion-imprinted systems commonly used SEM, FTIR, energy-dispersive X-ray spectroscopy, XRD, thermogravimetric analysis, BET analysis, and related elemental or thermal methods to verify functional binding environments in polymeric, silica-supported, carbon-supported, or hybrid matrices (Arbab-Zavar et al., 2011; Ebrahimzadeh et al., 2013; Fazelirad et al., 2014; Lesniewska et al., 2016; Liu et al., 2013; Zambrzycka et al., 2013). Magnetic and nanocomposite sorbents required additional confirmation of magnetic cores, imprinted shells, graphene-based components, mesoporous carbon, or metal–organic framework/zeolitic imidazolate framework support integration using vibrating-sample magnetometry, transmission electron microscopy, SEM, XRD, FTIR, BET analysis, and thermal analysis (Ansari and Masoum, 2018; Jalilian and Taheri, 2018; Li et al., 2021; Mamman et al., 2021; Phirisi et al., 2024; Sa-nguanprang et al., 2024; Sun et al., 2020; Wei et al., 2018). Quantitative characterization values were retained where they defined material structure rather than analytical performance. These included coating thicknesses of 77 nm and 20 nm in Fe₃O₄-based magnetic MIPs (Ciarlantini et al., 2025), magnetic susceptibility of 34.0 emu/g in dummy magnetic MIPs (Li et al., 2021), thermal stability up to 300 °C for Fe₃O₄@Cd-IIP (Panjali et al., 2016), a BET surface area of 279.6 m²/g for Fe₃O₄@SiO₂@chitosan–graphene oxide@MIP (Sun et al., 2020), and BET-derived surface areas of 238–279 m²/g for MIP-TiO₂ materials (Tonucci et al., 2020).

Figure 3 summarizes the section’s central framework by showing how programmable imprinted architectures become experimentally credible through characterization, condition-controlled operation, and matrix-compatible selective function.

Programmable synthesis routes and site accessibility

The synthesis routes used across the studies show how imprinting was programmed into different material formats. Bulk polymerization, precipitation polymerization, suspension polymerization, ultrasonic-assisted precipitation polymerization, thermal copolymerization, sol–gel processing, hydrothermal-assisted sol–gel synthesis, surface imprinting, emulsion templating, reversible addition–fragmentation chain-transfer polymerization, cyclic-voltammetric electrosynthesis, and activators regenerated by electron transfer atom-transfer radical polymerization-based surface imprinting were used to position recognition sites within bulk matrices, on accessible surfaces, in magnetic or core–shell particles, or on porous and hybrid supports (Abdullah et al., 2019; Abdullah et al., 2020; Abdullah et al., 2023; Akbarzadeh et al., 2022; Amran et al., 2024; Arbab-Zavar et al., 2011; Bhogal et al., 2021b; Chen et al., 2022; Darroudi et al., 2020; Guo et al., 2019; Khoddami and Shemirani, 2016; Li et al., 2018; Li et al., 2011; Liu et al., 2013; Mokhtari et al., 2022; Piryaei et al., 2021; Wang et al., 2016; Wei et al., 2018). These synthesis choices controlled not only the chemical identity of the binding sites but also their accessibility, matrix compatibility, and suitability for specific handling formats.

Surface, core–shell, and supported construction were especially important for translating molecular recognition into operationally usable formats. Silica particles, magnetic nanoparticles, mesoporous supports, carbon microspheres, polydopamine particles, cryogels, monoliths, straw substrates, hollow fibers, and stainless-steel solid-phase microextraction (SPME) fibers were used to place imprinted sites near accessible interfaces or within device-compatible platforms (Dinali et al., 2021; Du et al., 2014; Huang et al., 2019; Li et al., 2018; Luo et al., 2011; Pantanit et al., 2025; Pebdani et al., 2015; Piryaei et al., 2021; Rahimi et al., 2022; Wang et al., 2016; Wang et al., 2023; Wei et al., 2018; Yang et al., 2011). These architectures supported solid-phase extraction (SPE), microextraction by packed sorbent (MEPS), pipette-tip SPE, SPME, molecularly imprinted magnetic SPE, dispersive SPE, magnetic SPE, fixed-bed operation, cryogel adsorption, and sensor-coating workflows (Dinali et al., 2021; Li et al., 2018; Li et al., 2021; Mokhtari et al., 2022; Moradirad et al., 2024; Pantanit et al., 2025; Piryaei et al., 2021; Rahimi et al., 2022; Sartore et al., 2020; Shahar et al., 2018; Soleimani et al., 2018; Wang et al., 2016; Wei et al., 2018). Representative section-specific examples include 8 mg of core@mesoporous MIP in MEPS (Dinali et al., 2021), 2.0 mg of MIP-SBA-15 packed into a 100 μL pipette tip (Piryaei et al., 2021), hollow-fiber SPME for diclofenac (Pebdani et al., 2015), and fixed-bed formats for gas-phase H₂S adsorption (Moradirad et al., 2024). These examples show that programmability was expressed through both recognition chemistry and physical format.

Rational and computational design of recognition environments

Computational and theoretical tools provided an additional route to programmable recognition by guiding monomer selection, template–monomer interactions, solvent choice, and imprinting-configuration design. Density functional theory (DFT) was used in Cd(II), carvedilol, salicylic acid, dinotefuran, and palladium systems to evaluate functional monomers, template–monomer complexes, solvents, or imprinting configurations (Adauto et al., 2020; da Silva et al., 2017; Lai et al., 2019; Rahangdale et al., 2016; Silva et al., 2018). Molecular dynamics was also used in selected MIP–poly(high-internal-phase-emulsion) and chitosan-based systems to support interpretation of binding interactions within polymer environments (Abril et al., 2026; Akbarzadeh et al., 2022). Computational molecular modeling was used to optimize dummy-template polybrominated diphenyl ether systems, while virtual imprinting guided polychlorinated-biphenyl-targeted 4,4′-dihydroxydiphenyl ether-MIP design (Marc et al., 2018; Wang et al., 2025). These approaches connect molecular-level preorganization with experimentally realized selective interfaces.

The Pd-IIP study illustrates this design logic by comparing three imprinting configurations computationally, with calculated adsorption energies of −8.764, −13.978, and −3.587 eV for the three Pd-IIP configurations, respectively (Lai et al., 2019). Dinotefuran imprinting similarly used DFT to guide the selection of methacrylic acid as the functional monomer in a 1:4 template-to-monomer ratio with chloroform as the solvent before bulk polymerization and MEPS-high-performance liquid chromatography (HPLC) application (Silva et al., 2018). In these examples, computational design did not replace experimental verification; rather, it provided a rational basis for programming recognition-site chemistry before synthesis and workflow integration.

Condition control as a determinant of selective operation

The experimental conditions reported across the studies show that selective function depended on both the imprinted material and the environment in which it operated. Commonly optimized or controlled variables included pH, sorbent mass, polymer weight, sample volume, contact time, extraction time, desorption time, washing solvent, sample or loading solvent, eluent type and volume, eluent concentration, flow rate, temperature, reagent ratios, and draw–eject or aspiration/dispersion cycles. These variables determined whether recognition sites could capture and release the target under the chemical and physical constraints of complex samples.

In ion-imprinted systems, pH and eluent chemistry were particularly important because target capture and release depended on solution chemistry, coordination environment, and protonation or complexation equilibria. Representative conditions include Pb sorption at pH 6 in a flow-injection flame atomic absorption spectrometry system using 0.5 mL of 3 mol/L HNO₃ and a flow rate of 1.5 mL/min (Amran et al., 2024); Cr(III) retention at pH 9–10 with elution by 4 mol/L acetic acid (Lesniewska et al., 2016); and Cr(III)-IIP SPE at pH 3.0 using 50 mg sorbent, 15 min adsorption, 5 min desorption, and 3 mL of 4 mol/L HNO₃ as eluent (Roushani et al., 2018). Other representative operating conditions include magnetic Cd-IIP operation at pH 7 with 5 min sorption/elution and 4 mL of 2 mol/L HCl as eluent (Panjali et al., 2016), Cd(II)-IIP operation at pH 6.0 with 60 min equilibrium time and regeneration using 0.1 M HNO₃ (Wang et al., 2019), dynamic Ru(III) SPE at pH 8.0 ± 0.1 with elution by 2 mL of 0.3 mol/L thiourea in 0.3 mol/L HCl (Zambrzycka et al., 2013), and Cu(II)-IIP operation examined across pH 1–7 with a reported operating point at pH 5.0, 25 °C, and 10 min contact time (Zhang et al., 2017). Comparable condition control was reported for Tl, Cd, As, Se, Hg, Ag, La, and radioyttrium systems through optimization of pH, polymer mass, eluent nature, flow rate, sorption/desorption time, column operation, and batch variables (Arbab-Zavar et al., 2011; Darroudi et al., 2020; Li et al., 2011; Mafu et al., 2016; Rahman et al., 2017; Sarabadani et al., 2013; Taghizadeh and Shafizadeh, 2025; Wang et al., 2016; Zhang et al., 2014).

For organic molecules, pharmaceuticals, pesticides, endocrine disruptors, dyes, and food contaminants, optimization was shaped by matrix compatibility and extraction format. Core@mesoporous MIP-MEPS for pesticide analysis used 200 μL ultrapure water as the washing solvent, 150 μL acetonitrile as the eluent, 100 μL sample at pH 2.5, 8 mg adsorbent, and five draw–eject cycles (Dinali et al., 2021). Carvedilol pipette-tip MIP-SPE used 20 mg MIP for 500 μL urine at pH 12.5 before HPLC separation on a Chiralpak IA column with methanol:ethanol:water at 64:15:21 containing 0.3% diethylamine (da Silva et al., 2017). Amlodipine MEPS-capillary electrophoresis used 5.5 mg carboxymethyl-β-cyclodextrin in 125.0 mmol/L triethylamine at pH 6.0, with extraction optimized at pH 5 using 8.0 mg adsorbent and 150 μL methanol for elution (da Silva et al., 2024). Melatonin extraction by dispersive magnetic SPE used 10 mg magnetic MIP, pH 4.2, 14 min sorption, and elution with 0.180 mL methanol for 2.5 min (Dil et al., 2021). Other workflows optimized or controlled washing solvent, sample volume, eluent type and volume, aspiration/dispersion cycles, pH, extraction/desorption time, stirring rate, organic solvent volume, and solvent composition for estrogens, diclofenac, phenobarbital, mandelic acid, zearalenone, sarcosine, methoxyfenozide, cannabinoids, fluoroquinolones, phthalates, and patulin (de Oliveira et al., 2019; Mausia et al., 2011; Moein et al., 2015; Pebdani et al., 2015; Piryaei et al., 2021; Prieto et al., 2011; Rahimi et al., 2022; Sartore et al., 2020; Soleimani et al., 2018; Wei et al., 2018; Zhao et al., 2011).

Design-of-experiment approaches for multivariable control

Several studies moved beyond single-variable optimization by using formal design-of-experiment strategies. Response surface methodology, central composite design, face-centered design, Plackett–Burman screening, Taguchi methodology, univariate and multivariate experiments, and multi-objective response surface methodology were used to optimize sorbent composition, extraction variables, binding conditions, and gas-phase operating conditions (Andrade et al., 2023; Ansari and Masoum, 2018; Dil et al., 2021; Hayat et al., 2022; Mahabadi et al., 2025). This multivariable approach aligns closely with the manuscript’s focus because programmable imprinted soft materials often require simultaneous control of material composition, matrix exposure, and operating environment.

Fipronil dispersive SPE was optimized using response surface methodology with central composite and face-centered designs, with reported binding conditions of 18 min, 42 °C, pH 7, and 120 mg/L analyte concentration before HPLC analysis (Hayat et al., 2022). H₂S-imprinted materials were optimized by central composite design across adsorbent dosage of 0.1–1 g, temperature of 25 °C–80 °C, H₂S concentration of 10–1000 ppm, and flow rate of 40–100 mL/min (Mahabadi et al., 2025). A related H₂S-imprinted MIP system used response surface design and Taguchi optimization across 16 experiments, varying template amount, crosslinker, functional monomer, porogen ratio, and initiator amount, and was evaluated in a fixed bed of 400 mm height and 10 mm diameter (Moradirad et al., 2024). These examples show that condition control was not merely procedural; it was part of the programming strategy that linked material design to selective operation.

Matrix-specific workflows and instrumental integration

The reviewed studies paired imprinted materials with analytical workflows according to target chemistry, matrix complexity, and required readout. Metal and metalloid systems were commonly coupled with flame atomic absorption spectrometry, electrothermal atomic absorption spectrometry, inductively coupled plasma atomic emission spectrometry, atomic fluorescence spectroscopy, ultraviolet spectrophotometry, or gamma-ray spectrometry following SPE, magnetic SPE, ion-imprinted preconcentration, or radiochemical separation (Amran et al., 2024; Arbab-Zavar et al., 2011; Fazelirad et al., 2014; Jalilian and Taheri, 2018; Khoddami and Shemirani, 2016; Lesniewska et al., 2016; Liu et al., 2013; Panjali et al., 2015; Panjali et al., 2016; Roushani et al., 2018; Sarabadani et al., 2013; Zambrzycka et al., 2013; Zhang et al., 2014). Organic compounds, drugs, pesticides, endocrine disruptors, dyes, and bioactive molecules were commonly analyzed by HPLC, HPLC with ultraviolet detection, HPLC with diode-array or fluorometric detection, liquid chromatography–tandem mass spectrometry (LC-MS/MS), liquid chromatography–time-of-flight mass spectrometry, hydrophilic interaction chromatography–tandem mass spectrometry, gas chromatography–mass spectrometry, capillary electrophoresis, fluorescence spectroscopy, diffuse reflectance spectrophotometry, fiber-optic linear-array spectrophotometry, flowing atmospheric-pressure afterglow mass spectrometry, or spectrophotometry after MEPS, molecularly imprinted SPE, pipette-tip SPE, SPME, molecularly imprinted magnetic SPE, dispersive SPE, magnetic SPE, or MIP-based MEPS (Ameli and Alizadeh, 2012; Andrade et al., 2023; Ansari and Masoum, 2018; da Silva et al., 2017; da Silva et al., 2024; Dil et al., 2021; Du et al., 2014; Guc et al., 2021; Ji et al., 2013; Li et al., 2018; Moein et al., 2015; Mortari et al., 2021; Pajewska-Szmyt et al., 2020; Pantanit et al., 2025; Pebdani et al., 2015; Phirisi et al., 2024; Prieto et al., 2011; Rahimi et al., 2022; Sartore et al., 2020; Soledad-Rodríguez et al., 2017; Soleimani et al., 2018; Wang et al., 2023; Wei et al., 2018; Zhao et al., 2011). In these examples, the imprinted material supplied selective recognition, while the extraction format and instrument made that recognition compatible with the matrix.

Biological matrices such as urine, plasma, breast milk, and serum were addressed through MEPS, pipette-tip SPE, SPME, magnetic extraction, LC-MS/MS-compatible formats, and related enrichment workflows (da Silva et al., 2017; da Silva et al., 2024; Daryanavard et al., 2013; Moein et al., 2015; Pajewska-Szmyt et al., 2020; Sartore et al., 2020; Wang et al., 2023). Food and agricultural matrices such as honey, milk, apple juice, fruit juice, vegetables, wine, and corn were treated using MIP-MEPS, magnetic SPE, HPLC-based workflows, LC-MS/MS, hydrophilic interaction chromatography–tandem mass spectrometry, and gas chromatography–mass spectrometry (Andrade et al., 2023; Chen et al., 2022; Dinali et al., 2021; Ji et al., 2013; Orachorn and Bunkoed, 2022; Pantanit et al., 2025; Phirisi et al., 2024; Soledad-Rodríguez et al., 2017; Wei et al., 2018; Zhao et al., 2011). Environmental waters and wastewater were processed by molecularly imprinted SPE, magnetic SPE, ion-imprinted SPE, batch adsorption, or column procedures before chromatographic, spectrometric, or atomic-spectrometric detection (Amran et al., 2024; Herrero-Hernández et al., 2013; Lesniewska et al., 2016; Liu et al., 2013; Mafu et al., 2016; Panjali et al., 2016; Prieto et al., 2011; Roushani et al., 2018; Wang et al., 2021; Zambrzycka et al., 2013; Zhang et al., 2014). Radiochemical separation required a distinct operational design: radioyttrium was separated after irradiation of a natural SrCO₃ target with a 15 MeV proton beam at 30 μA for 12 min, followed by optimization of pH, sorbent mass, and eluent nature and verification by inductively coupled plasma atomic emission spectrometry and gamma-ray spectrometry (Sarabadani et al., 2013).

Luminescence, spectroscopic, and sensor-oriented workflows further extended the operational range of imprinted soft-material platforms. Chemiluminescence systems for clopidogrel included Ru(phen)₃ ²⁺–Ce(IV), KMnO₄–H₂SO₄, and luminol–H₂O₂ (Mokhtari et al., 2022). Sonochemiluminescence for bentazone used a low-power humidifier as a pulsed ultrasound generator during MIP–poly(high-internal-phase-emulsion) preconcentration (Akbarzadeh et al., 2022). Diffuse reflectance spectrophotometry enabled direct solid-phase measurement of Acid Green 16 after adsorption assays (Mortari et al., 2021), while fiber-optic linear-array spectrophotometry was paired with hollow-fiber MIP-SPME for diclofenac (Pebdani et al., 2015). Chemiresistive sensing was demonstrated using silver nanoparticle-core MIP films deposited on interdigitated array electrodes by drop casting and dip coating for gas-phase linalool recognition (Shahar et al., 2018). Electrical MIP-based sensor formats further illustrate how imprinted recognition layers can be coupled to direct transduction. A paracetamol-selective molecularly imprinted copolymer was prepared by thermal free-radical bulk polymerization using methacrylic acid and styrene as functional monomers, EGDMA as the crosslinker, AIBN as the initiator, and dimethyl sulfoxide as the porogenic solvent; FTIR was used for polymer characterization, and an inductance–capacitance–resistance meter measured electrical parameters, including inductance and resistance (Shafique et al., 2024). A moxifloxacin MIP sensor was similarly synthesized by thermal free-radical bulk polymerization using methacrylic acid, EGDMA, AIBN, and dimethyl sulfoxide, then integrated with interdigital electrodes and evaluated through inductance–capacitance–resistance-meter-based measurement of concentration-dependent electrical properties (Memoona et al., 2024). These workflows show that imprinted platforms can operate as sorbents, sample-preparation media, separation phases, assay interfaces, and sensor recognition layers within the boundaries established by their characterization and operating conditions.

Table 3 organizes these patterns by linking each design-to-function category with the relevant programmable logic, section-specific experimental data, operational context, and representative references.

TABLE 3. Characterization and experimental-condition patterns supporting programmable imprinted soft materials in complex media.

Responsive, hybrid, and externally controlled operation

The most direct examples of programmable operation involved materials whose recognition or release behavior could be externally or structurally regulated. Thermosensitive MIPs for bisphenol A were evaluated below and above the lower critical solution temperature of 33 °C, particularly at 20 °C and 50 °C (Dong et al., 2014), while thermally responsive mesoporous silica nanosphere MIPs for dichlorophen were characterized by SEM, transmission electron microscopy, FTIR, and XRD (Guo et al., 2019). Photonic and magnetic dual-responsive Fe₃O₄@MIPs used 365 nm irradiation and subsequent visible-light exposure to regulate caffeine release and rebinding (Xu et al., 2013). Magnetic architectures were repeatedly used to simplify phase separation from aqueous, food, beverage, biological-fluid, and environmental matrices (Ansari and Masoum, 2018; Dil et al., 2021; Jalilian and Taheri, 2018; Li et al., 2021; Pajewska-Szmyt et al., 2020; Phirisi et al., 2024; Sa-nguanprang et al., 2024; Sun et al., 2020; Wei et al., 2018).

Hybrid recognition systems extended imprinting by incorporating additional soft, biofunctional, affinity-based, or nanostructured interfaces. Aptamer–MIP architectures on zeolitic imidazolate framework-67 used gold-nanoparticle-mediated immobilization of thiol-modified aptamers (Geng et al., 2024). Boronic-acid-modified MIPs were coupled with graphene oxide–silver–glutathione materials for ellagic acid enrichment (Gao et al., 2024). Bowl-shaped polydopamine-based MIPs used base-complementary pairing between 2′-deoxyadenosine and a pyrimidine monomer (Wang et al., 2023). A parathion MIP sorbent assay used a noncovalently imprinted polymer as the capture reagent and a p-aminoparathion–horseradish peroxidase conjugate as the enzyme label, with optimized assay conditions of 10 ng/mL p-aminoparathion–horseradish peroxidase conjugate and 10 mg/mL polymer (Tang and Xiang, 2010). Gas-phase workflows broadened the operating space further: nicotine surface-imprinted monoliths were evaluated as gas-chromatographic stationary phases (Huang et al., 2019), H₂S-imprinted materials were tested under controlled gas concentration, temperature, flow, and fixed-bed conditions (Mahabadi et al., 2025), and silver nanoparticle-core MIP chemiresistor films were used for volatile recognition (Shahar et al., 2018). These systems align closely with the concept of programmable imprinted soft materials because recognition was coupled to magnetic handling, temperature, light, gas flow, enzyme-labeled assay interfaces, affinity-based hybrid recognition, or electrical transduction.

Taken together, the characterization and experimental-condition evidence supports a design-to-function framework for programmable imprinted soft materials in complex media. The reviewed studies show that selective function was not established by imprinting alone. It depended on verified material structure, accessible recognition-site placement, computationally or experimentally guided design, matrix-specific condition control, and compatible analytical or operational workflows. Characterization confirmed whether the intended architecture existed; optimization defined the conditions under which recognition could occur; and workflow integration translated selective binding into practical extraction, separation, sensing, assay, gas-phase, radiochemical, or analytical operation. The addition of MIP-based electrical sensor examples further reinforces the role of imprinted soft-material interfaces as transduction-ready recognition layers. This relationship among programmed architecture, experimental condition, and matrix-specific function is the central contribution of this section.

Adsorption, selectivity, and analytical performance

The adsorption, selectivity, and analytical performance of imprinted soft materials provide the quantitative basis for evaluating whether these systems deliver programmable selective function in complex media. Here, performance is not defined simply by high uptake, but by the coordinated expression of target recognition, selective enrichment, interference resistance, tunable uptake or release, repeated usability, and, where sensor formats are used, reversible transduction. The strongest evidence comes from studies that directly compare imprinted materials with non-imprinted polymers or conventional sorbents, because such comparisons distinguish imprinting-derived recognition from nonspecific adsorption. These advantages were reported for several representative systems, including Cd(II), Co(II), H₂S, and Cd(II)-magnetic imprinted materials, where imprinted architectures showed higher capacity, improved selectivity, or stronger operational performance than their comparators (Buhani et al., 2010; Khoddami and Shemirani, 2016; Mahabadi et al., 2025). Carbamazepine-selective bio-based MIPs further demonstrated selective adsorption in pharmaceutical-spiked wastewater (Savigni et al., 2025). Figure 4 frames this section as a performance-validation pathway, linking complex media, selective imprinted interfaces, measurable outcomes, and evidence strength.

Table 4 complements the narrative by grouping representative outcomes according to the main evidence axes used to evaluate programmable selective function in complex media.

TABLE 4. Adsorption, selectivity, and analytical-performance outcomes demonstrating programmable imprinted function in complex media.

Selective enrichment, sensing, and analytical sensitivity

Imprinted materials often function as selective enrichment or sensing interfaces that concentrate or recognize trace analytes before instrumental or electrical detection. This role is especially important in complex media, where target compounds occur at low concentrations and coexist with ions, organic matter, metabolites, food constituents, or matrix interferents. Representative metal-ion methods achieved low limits of detection (LODs) for Pb²⁺, As³⁺, Cd²⁺/Pb²⁺, Cr(VI), Ru(III), and mercury species, including LODs of 0.59 μg/L for Pb²⁺, 0.65 μg/L for As³⁺, 0.25–0.39 μg/L for Pb²⁺/Cd²⁺ in a dual-metal system, 0.81 μg/L for Cr(VI), 0.26 ng/mL for Ru(III), and 0.015–0.02 μg/L for mercury species (Abdullah et al., 2019; Abdullah et al., 2020; Abdullah et al., 2023; Kumar et al., 2020; Zambrzycka et al., 2013; Zhang et al., 2014).

Organic and pharmaceutical targets showed comparable sensitivity gains when imprinting was integrated with preconcentration. MIP–poly(high-internal-phase-emulsion) preconcentration lowered the bentazone LOD from 8.4 × 10⁻⁸ to 8.8 × 10⁻¹⁰ mol/L, while a clopidogrel MIP–poly(high-internal-phase-emulsion) system increased chemiluminescence sensitivity and achieved an LOD of 3.0 × 10⁻¹⁰ mol/L (Akbarzadeh et al., 2022). Other representative systems reached ng/mL, ng/L, or low μg/L detection for parabens, cannabinoids, carbamates, mandelic acid, and phthalic acid esters (Pajewska-Szmyt et al., 2020; Sartore et al., 2020; Soleimani et al., 2018; Wang et al., 2021; Wei et al., 2018).

Electrical MIP sensor formats extend this performance logic by converting selective recognition into reversible electrical responses. A paracetamol-imprinted copolymer sensor showed concentration-dependent response, with LODs of 12 mg/L by inductance and 5.33 mg/L by resistance, together with selectivity against ascorbic acid and uric acid, reproducibility, and reversible response (Shafique et al., 2024). Similarly, a moxifloxacin MIP sensor based on interdigital electrodes achieved LODs of 240 ppb for series resistance, 63 ppb for parallel resistance, 48 ppb for series inductance, and 8 ppb for parallel inductance, while retaining selectivity against uric acid, ascorbic acid, and paracetamol, with reversible and reproducible response (Memoona et al., 2024). A non-imprinted Bi₂O₃@porous carbon electrochemical sensor also provides a useful practical benchmark for Hg²⁺ detection, with a detection limit of 98 nM, sensitivity of 0.101 μA μM^-1, R ² = 0.996, relative standard deviation (RSD) of 1.19%, 89.6% signal retention over 30 days, and 99.12% ± 3.45% recovery in canal water; however, this system should be interpreted as an electrochemical deployment comparator rather than evidence of imprinting-derived selectivity (Mushtaq et al., 2026). Quantitative validation in many cited examples was supported by recoveries near 80%–110% and RSDs below 7%, although broader precision ranges were reported in some cases, including 12%–18% for phenol and phenoxyacid herbicide monitoring and up to 14.6% interday RSD for aminoglycosides (Herrero-Hernández et al., 2013; Ji et al., 2013). These outcomes show that the analytical value of imprinted soft materials lies in combining recognition with measurable enrichment or transduction, rather than in passive extraction alone.

Adsorption capacity and removal efficiency

Capacity and removal data indicate whether selective recognition translates into functionally meaningful uptake. Because target classes and reporting units vary widely, these values are most informative within target-specific comparisons rather than as direct cross-system rankings. Among inorganic targets, high-capacity examples included 169.49 mg/g for Cr(VI), 125 mg/g for La³⁺, 96 mg/g for Cu²⁺ removal from concentrated zinc sulfate solutions, 83.89 mg/g for Cd(II)-imprinted silica, 77.6 ± 0.6 mg/g for Ag(I), and 74.65 mg/g for Cr(III) (Buhani et al., 2010; Kumar et al., 2020; Roushani et al., 2018; Taghizadeh and Shafizadeh, 2025; Wang et al., 2016; Zhang et al., 2017). Organic pollutants and bioactive compounds also showed substantial uptake, including 109.1 mg/g for melatonin, 108.35 mg/g for dichlorophen, 85 mg/g for diethyl phthalate, 79.36 mg/g for sotalol, 65.1 mg/g for ochratoxin A, 54.4 mg/g for fenthion, and 47 mg/g for fipronil (Ansari and Masoum, 2018; Dil et al., 2021; Geng et al., 2024; Geng et al., 2025; Guo et al., 2019; Hayat et al., 2022). Particularly high uptake was reported for the related zirconium bio-based MIP-202 adsorbent toward short-chain per- and polyfluoroalkyl substances, with adsorption capacities of 620.5, 603.2, 573.9, and 525.7 mg/g for perfluorobutanesulfonate, perfluoropentanesulfonate, perfluorobutanoate, and perfluoropentanoate, respectively (Borui et al., 2024). Removal efficiencies further support the functional relevance of these materials, including more than 90% patulin removal, 100% 17β-estradiol removal through combined adsorption and photocatalysis, 99.9% Cu²⁺ removal with residual copper reduced to 35 μg/L, and 94.3% H₂S adsorption efficiency compared with 9.9% for the non-imprinted control (Mahabadi et al., 2025; Sun et al., 2020; Tonucci et al., 2020; Zhang et al., 2017).

Selectivity in competitive and complex media

Selectivity is the central performance criterion for imprinted soft materials because these systems are intended to operate in media containing competing ions, structurally related molecules, and matrix interferents. Ion-imprinted systems provided some of the clearest evidence of selective function. Pb²⁺ selectivity coefficients reached 19.98, 42.15, and 42.96 against Cu²⁺, Cd²⁺, and Zn²⁺; Cd(II) selectivity coefficients reached 99.798, 93.045, and 86.617 against Ni(II), Cu(II), and Zn(II); Co(II) separation factors reached 41.17, 79.74, and 56.48 against Pb(II), Ni(II), and Cd(II); and Hg²⁺ selectivity factors ranged from 19 to 34 over other mercury forms (Ambarsari et al., 2023; Buhani et al., 2010; Khoddami and Shemirani, 2016; Zhang et al., 2014). Anti-interference performance was also reported for As³⁺, Pb²⁺/Cd²⁺, arsenic/selenium, Cr(III), and additional metal-ion systems under competitive conditions (Abdullah et al., 2020; Abdullah et al., 2023; Mafu et al., 2016; Roushani et al., 2018).

MIPs demonstrated selectivity through imprinting factors, cross-reactivity profiles, and preferential adsorption over structural analogues. Representative values included imprinting factors of 5.48 for ochratoxin A, 3.01 for ellagic acid, 2.94 for fenthion, and 1.30 for 2′-deoxyadenosine (Gao et al., 2024; Geng et al., 2024; Geng et al., 2025; Wang et al., 2023). Zearalenone enrichment showed 96.9% selectivity, with cross-reactivity of 1%–20% toward several related toxins and 61% toward fumonisin B₁ (Mausia et al., 2011). Other selective systems included nicotine, with a selectivity factor of 10.15; testosterone, with a high-affinity association constant of 3.3 × 10⁷ M^-1; theophylline recognition over caffeine; selective fluoroquinolone retention despite structurally diverse competitors; preferential patulin adsorption over 5-hydroxymethylfurfural; and MIP-based electrical/electrochemical discrimination of paracetamol or moxifloxacin from common electroactive interferents (Bui and Haupt, 2011; Huang et al., 2019; Luo et al., 2011; Prieto et al., 2011; Zhao et al., 2011; Shafique et al., 2024; Memoona et al., 2024).

Programmable uptake, release, and responsive performance

The manuscript’s central concept of programmability is most directly supported by systems in which binding, release, recovery, or signal response can be tuned by formulation, architecture, or external stimulus. Phosphorus sorption capacity increased from approximately 11–28 mg PO₄ ^3-–P g^-1 by changing the template-to-monomer ratio, showing that imprinting formulation can regulate uptake (Ritt et al., 2019). Thermosensitive bisphenol A materials showed temperature-dependent adsorption and release behavior (Dong et al., 2014), while pH-responsive diethyl phthalate sorbents reached 85 mg/g uptake at pH 5 and 92% desorption at pH 8 (Wolska et al., 2020). Photonic and magnetic caffeine MIPs further demonstrated controlled capture and release, with 62.5% of receptor-bound caffeine released under 365 nm irradiation and 93.6% rebound under visible light (Xu et al., 2013). MIP-based electrical sensors add another programmable-function mode, in which molecular recognition is coupled to reversible electrical readouts, as shown for paracetamol and moxifloxacin detection (Shafique et al., 2024; Memoona et al., 2024). Hybrid recognition systems also supported programmable function by combining imprinted sites with additional recognition or transduction elements. For example, an ochratoxin A aptamer/MIP/zeolitic imidazolate framework-67 adsorbent showed an adsorption capacity of 65.1 mg/g and an imprinting factor of 5.48, while other imprinted or hybrid systems used magnetic, carbon-based, metal–organic framework/zeolitic imidazolate framework, or photocatalytic components to support extraction, recovery, or removal performance (Geng et al., 2024; Orachorn and Bunkoed, 2022; Tonucci et al., 2020). These examples show that programmability is most convincingly demonstrated when material design produces measurable changes in uptake, release, recovery, regeneration, or signal response.

Kinetic, isotherm, and preconcentration behavior

Kinetic and isotherm models help explain why imprinted materials perform effectively by describing binding-site accessibility, adsorption rate, surface or interfacial adsorption behavior, and affinity heterogeneity. Langmuir isotherm behavior and pseudo-second-order kinetics were repeatedly reported for Pb²⁺, As³⁺, Pb²⁺/Cd²⁺, Cd²⁺, Cr(VI), Ag(I), Hg²⁺, H₂S, and dibenzothiophene systems (Abdullah et al., 2019; Abdullah et al., 2020; Abdullah et al., 2023; Kumar et al., 2020; Li et al., 2011; Mahabadi et al., 2025; Taghizadeh and Shafizadeh, 2025; Yang et al., 2011; Zhang et al., 2014). Other systems followed Freundlich, Langmuir–Freundlich, pseudo-first-order, first-order, or second-order reversible models depending on the target and material, including fipronil, bisphenol A, patulin, polychlorinated biphenyls, salicylic acid, and Hg(II)-imprinted monoliths (Dong et al., 2014; Hayat et al., 2022; Rahangdale et al., 2016; Rahangdale and Kumar, 2021; Rahman et al., 2017; Sun et al., 2020; Wang et al., 2025).

Preconcentration metrics further show that imprinted materials can act as active analytical interfaces. Pb²⁺ flow-injection flame atomic absorption spectrometry achieved enrichment above 20-fold, a concentration efficiency of 11.65 min^-1, and a consumptive index of 4.5 mL (Amran et al., 2024); bentazone preconcentration factors reached 113–131 (Akbarzadeh et al., 2022); Cr(III)-IIP provided a preconcentration factor of 100 (Roushani et al., 2018); and ampicillin molecularly imprinted SPE achieved a preconcentration factor of 20, with recoveries above 98% and RSD below 7% (Soledad-Rodríguez et al., 2017). Together, these findings support the interpretation of imprinted soft materials as tunable enrichment and recognition systems rather than inert adsorbents.

Operational durability and reusability

Reusable performance determines whether selective function can be maintained beyond a single extraction, adsorption, or sensing event. The strongest evidence came from systems with explicit cycle or extraction counts. A sulfonylurea herbicide MEPS device was reused more than 180 times without loss of extraction capability, and a cannabinoid MEPS platform operated for more than 90 consecutive extractions without carryover (Andrade et al., 2023; Sartore et al., 2020). Other materials retained function over 20 cycles for dual endocrine-disruptor extraction, 15 cycles for Hg(II)-imprinted monoliths, 14 cycles for phenylurea herbicide extraction, 10 cycles for La³⁺, mercury-speciation, and multi-template pharmaceutical adsorbents, eight cycles for the related MIP-202 per- and polyfluoroalkyl substance adsorbent and dibenzothiophene systems, and seven cycles for melatonin, Co(II), microcystin-LR, and dibutyl phthalate systems (Bhogal et al., 2021b; Borui et al., 2024; Dil et al., 2021; Khoddami and Shemirani, 2016; Krupadam et al., 2012; Li et al., 2024; Pantanit et al., 2025; Rahman et al., 2017; Venkatesh et al., 2014; Wang et al., 2016; Yang et al., 2012; Zhang et al., 2014). More specific retention values included 97.6% catalytic performance after five uses for flavin-imprinted magnetic polymers, 91.14% capacity retention after four adsorption–desorption cycles for 2′-deoxyadenosine, and 74% capacity retention after six Cd(II)-IIP reuse cycles (Liu et al., 2022; Wang et al., 2019; Wang et al., 2023). Electrical/electrochemical sensing examples also reported reversible and reproducible response for imprinted paracetamol and moxifloxacin sensors, while the non-imprinted Bi₂O₃@porous carbon Hg²⁺ sensor provided a practical stability benchmark with 89.6% signal retention over 30 days (Shafique et al., 2024; Mushtaq et al., 2026; Memoona et al., 2024). Where studies reported only qualitative stability, recyclability, or selectivity without cycle numbers or retained-capacity values, those claims are best interpreted as supportive but less definitive operational evidence (Özkahraman et al., 2011; Piryaei et al., 2021; Rahiminezhad et al., 2010; Silva et al., 2018).

Real-sample application and review-level significance

Real-sample validation is the most direct test of whether imprinted soft materials can deliver selective function in complex media. In this section, programmability refers not only to template-defined recognition, but also to matrix-tolerant extraction, selective removal, controlled uptake and release, magnetic retrieval, regeneration, speciation, recovery, and signal transduction under conditions where competing ions, endogenous biomolecules, organic matter, food components, or industrial co-contaminants may interfere. The studies reviewed here move imprinted materials beyond idealized binding systems and show where encoded selectivity remains useful in environmental, food, biomedical, industrial, and resource-oriented settings. The relationship between programmable material features, complex-media validation, selective function, and review-level significance is summarized conceptually in Figure 5 and analytically in Table 5.

TABLE 5. Real-sample validation patterns for programmable imprinted soft materials in complex media.

Environmental monitoring and water remediation

Environmental waters provide the broadest validation space for programmable imprinting because they contain competing species, matrix components, and multiple pollutant classes. Across the reviewed literature, imprinted sorbents were applied to river water, groundwater, bottled water, seawater, wastewater, municipal sewage, road runoff, industrial aqueous samples, vineyard-region surface water and groundwater, eutrophic waters, and extracts of particulate matter with aerodynamic diameter ≤2.5 μm (Adauto et al., 2020; Amran et al., 2024; Bhogal et al., 2021b; Chen et al., 2022; Dong et al., 2014; Herrero-Hernández et al., 2013; Panjali et al., 2016; Rahman et al., 2017; Ritt et al., 2019; Zambrzycka et al., 2013; Zhang et al., 2014). IIPs were especially prominent for trace-metal determination, speciation, recovery, and removal. Representative examples include Pb(II) sorbents validated using certified reference materials and spike-addition methods (Abdullah et al., 2019), Cr(III)-imprinted polymers applied to wastewater and reference material RES 25.2 (Lesniewska et al., 2016), and Cd(II)-imprinted magnetic systems validated with Seronorm LOT NO2525 and spiked real samples before application to human urine, food, river water, and wastewater (Panjali et al., 2015). These studies show that the central value of imprinting in environmental matrices is not merely adsorption, but selective recognition in the presence of competing ions and complex backgrounds.

The same logic extends to organic contaminants and nutrients. Monitoring studies demonstrated direct environmental relevance, including bisphenol A detection in more than 50% of La Rioja vineyard-region surface-water and groundwater samples at concentrations up to 0.72 μg/L, with 2,4-dichlorophenoxyacetic acid exceeding 0.1 μg/L in two cases (Herrero-Hernández et al., 2013). Carbamate-selective magnetic MIPs detected carbaryl at 2.4 ng/L in Yitong River water, illustrating how magnetic extraction can streamline pesticide determination in natural waters while reducing matrix effects (Wang et al., 2021). Municipal wastewater was used to verify fluoroquinolone-selective MIPs in a complex wastewater matrix (Prieto et al., 2011), and bio-based MIPs selectively removed carbamazepine from pharmaceutical-spiked real wastewater over diclofenac and ibuprofen (Savigni et al., 2025). These examples establish complex water as a proving ground for matrix-tolerant selectivity rather than a passive application setting.

Remediation-oriented studies further broaden the significance of imprinted materials from analysis to selective function in treatment contexts. Phosphorus-imprinted polymers were designed for low-concentration phosphorus recovery from eutrophic surface waters, linking water treatment with reuse of a nonrenewable agricultural resource (Ritt et al., 2019). A microcystin-LR MIP was estimated to treat 3,640 L of water containing 1.0 μg/L toxin per gram of sorbent at an estimated cost of $1.5, while outperforming powdered activated carbon and XAD resin (Krupadam et al., 2012). MIP-TiO₂ achieved complete 17β-estradiol removal through combined adsorption and photocatalysis and remained reusable over five cycles (Tonucci et al., 2020), while a multi-template pharmaceutical MIP outperformed XAD resin and powdered activated carbon and remained effective over 10 adsorption–desorption cycles (Venkatesh et al., 2014). Together, these studies demonstrate that programmable imprinting can support selective detection, cleanup, recovery, and reusable remediation in chemically crowded aquatic systems.

Food, beverage, and agricultural matrices

Food and agricultural matrices test whether imprinted materials can selectively isolate or remove target contaminants without being overwhelmed by matrix-derived coextractives and interferents. The reviewed studies applied imprinted sorbents to milk, honey, wine, apple juice, fruit juice, soft drinks, yogurt, vegetables, pork, fish, corn, cow milk, and other agricultural products. The targets included pesticide residues, antibiotics, herbicides, bisphenols, phthalates, mycotoxins, clenbuterol, fipronil, patulin, nonsteroidal anti-inflammatory drugs, and phenylurea herbicides. These applications are important because they connect selective recognition to food-safety monitoring, matrix-compatible cleanup, and practical sample preparation.

Several examples show how real food matrices reveal the practical advantages of programmable imprinting. A sulfonylurea herbicide MIP–MEPS device was applied to ground corn and reused more than 180 times using only 1.0 mg of MIP, while showing higher uptake than non-imprinted polymers and commercial sorbents (Andrade et al., 2023). In Cabernet Sauvignon wine, chitosan-based MIPs removed 15%–40% of m-cresol while minimally affecting key phenolic parameters, demonstrating selective removal of a smoke-taint marker without substantially disrupting important wine-quality attributes (Abril et al., 2026). Patulin-focused sorbents further illustrate food-matrix relevance: one magnetic MIP removed more than 90% of patulin from apple juice after 24 h without negatively affecting final juice-quality parameters (Sun et al., 2020), while a patulin-imprinted silica material achieved recoveries above 90% in spiked apple juice and outperformed traditional liquid–liquid extraction, which achieved 75% recovery (Zhao et al., 2011).

Residue-monitoring studies also show how imprinting can support regulatory or safety-focused analysis in complex foods. Aminoglycoside-selective MIPs enabled simultaneous trace-antibiotic analysis in honey with negligible adsorption of non-aminoglycosides (Ji et al., 2013). Ampicillin-imprinted molecularly imprinted SPE achieved recoveries above 98% in cow milk and provided sufficient preconcentration to assess the European Union maximum residue limit of 4 μg/kg (Soledad-Rodríguez et al., 2017). Additional representative applications include nonsteroidal anti-inflammatory drug extraction from milk with six-cycle sorbent reuse (Orachorn and Bunkoed, 2022), phenylurea herbicide extraction from fruit juices with stability over 14 cycles (Pantanit et al., 2025), phenylurea herbicide extraction from vegetable samples with four-cycle reuse (Sa-nguanprang et al., 2024), and phthalate determination in water, milk, and wine with recoveries of 80.2%–103% (Wei et al., 2018). In this application domain, imprinted soft materials are most compelling when they combine selective cleanup, reusability, miniaturization, or magnetic handling with credible performance in real food and agricultural matrices.

Biomedical, clinical, forensic, and pharmaceutical relevance

Biological and pharmaceutical matrices provide a different challenge: targets must be recognized amid endogenous matrix components and structurally related compounds. The reviewed studies applied imprinted materials to urine, plasma, serum, breast milk, physiological fluids, human hair, fish meat, artificial gastric and intestinal fluids, and patient fecal samples. The targets included therapeutic drugs, enantiomers, hormones, metabolites, biomarkers, toxic metals, parabens, cannabinoids, mercury species, clopidogrel, paclitaxel, and 2′-deoxyadenosine. Here, the significance of imprinting lies in selective bioanalysis, cleanup, exposure monitoring, forensic extraction, synthetic receptor behavior, and controlled release in complex biological or pharmaceutical media.

Clinical and pharmaceutical examples illustrate this function clearly. MIP-based approaches were applied to sotalol in spiked human urine and plasma (Ansari and Masoum, 2018), carvedilol in a preliminary clinical urinary-excretion study (da Silva et al., 2017), and amlodipine enantiomers in urine after volunteer dosing, where 688.2 ng/mL of (+)-(R)-amlodipine and 689.8 ng/mL of (−)-(S)-amlodipine were detected, corresponding to approximately 14.8% elimination of the ingested racemate for both enantiomers (da Silva et al., 2024). MIP cartridges for local anesthetics remained stable for 100 extractions in plasma and urine workflows, demonstrating practical durability for repeated clinical sample preparation (Daryanavard et al., 2013). Restricted-access MIPs for estrogens were validated using urine from a pregnant volunteer, showing the importance of matrix-interference reduction in hormone monitoring (de Oliveira et al., 2019).

Other biomedical applications extend the role of imprinted materials from drug analysis to diagnostics, exposure monitoring, forensic workflows, and delivery. Dummy MIP–MEPS was applied to plasma and urine for sarcosine, a prostate-cancer biomarker, with recoveries of 87% and 89% (Moein et al., 2015). Automated MIP–MEPS for cannabinoids in human urine enabled more than 90 consecutive extractions without carryover, supporting robust clinical or forensic analysis (Sartore et al., 2020). Molecularly imprinted MEPS for mandelic acid required only 5 min of sample preparation, indicating practical relevance for occupational biomonitoring of exposure to ethylbenzene or styrene (Soleimani et al., 2018). Testosterone-imprinted receptors were compared with commercial antibody-based kits as stable synthetic recognition elements (Bui and Haupt, 2011), and MIP–poly(high-internal-phase-emulsion) enabled the first chemiluminescence-based determination of clopidogrel in pharmaceuticals and blood serum, with sensitivity increased up to 42-fold (Mokhtari et al., 2022). Paclitaxel dummy-template MIPs released approximately 45% of the drug across artificial gastric juice, artificial intestinal fluid, and patient fecal samples, supporting intestinal delivery relevance (Tan et al., 2021). Mercury IIPs also enabled discrimination between inorganic and organic mercury species in complex matrices such as human hair and fish meat while remaining stable over 10 cycles (Zhang et al., 2014). Across these examples, programmed recognition becomes valuable when biological matrix complexity would otherwise obscure, retain, or interfere with the target.

Industrial purification, gas treatment, and resource recovery

Industrial and process-oriented applications show that imprinted soft materials can be programmed for selective purification, not only analytical extraction. H₂S-imprinted materials showed selective removal in CO₂/H₂S/CH₄ gas mixtures, outperformed non-imprinted controls, and were positioned for industrial gas-stream purification and cleaner energy processing (Mahabadi et al., 2025). Fuel-related studies applied imprinting to pyridine removal from model fuel and dibenzothiophene recognition or removal in gasoline and complex mixtures, supporting denitrogenation and desulfurization strategies relevant to fuel upgrading and cleaner-fuel processing (Mathidala and Ogunlaja, 2019; Yang et al., 2012; Yang et al., 2011). In hydrometallurgical purification, Cu(II)-IIPs removed 99.9% of copper from concentrated ZnSO₄ solution and reduced residual Cu²⁺ to 35 μg/L, below the 100 μg/L threshold required for zinc processes (Zhang et al., 2017). These studies demonstrate selective function in process media that are chemically demanding and application-specific.

Resource recovery further expands the practical scope of imprinting. Phosphorus-imprinted polymers linked the treatment of eutrophic waters with the recovery of a nonrenewable agricultural resource (Ritt et al., 2019). Silver-imprinted magnetic materials enabled selective Ag(I) separation from aqueous media even in the presence of competing ions (Jalilian and Taheri, 2018), while straw-supported La³⁺-IIPs maintained their initial binding affinity over 10 cycles, supporting selective lanthanum recovery (Wang et al., 2016). In radiochemistry, ion-imprinted nanoparticles separated radioyttrium from a strontium target with 99.5% recovery and residual stable strontium of 4.5 μg/mL, supporting radionuclide production for medical and research applications (Sarabadani et al., 2013). These studies shift the interpretation of imprinted soft materials from contaminant capture alone toward selective harvesting of valuable, strategic, or process-critical targets.

Programmable architectures, comparative benchmarks, and field-level boundaries

The most direct connection to the manuscript’s central theme appears in studies where material architecture or responsiveness is linked to a programmed function in complex media. Dummy-template and substitute-template imprinting reduced template-leakage concerns, avoided toxicity or cost limitations, or improved recognition when direct imprinting was problematic, as shown for organophosphorus pesticides, phthalates, polybrominated diphenyl ethers, salicylic acid, paclitaxel, and patulin (Danmaraya et al., 2025; Li et al., 2024; Marc et al., 2018; Rahangdale et al., 2016; Tan et al., 2021; Zhao et al., 2011). Stimuli-responsive systems added programmed control over uptake and release, including temperature-regulated bisphenol A removal in seawater and yogurt (Dong et al., 2014), thermally responsive dichlorophen separation for wastewater treatment (Guo et al., 2019), pH-controlled diethyl phthalate sorption and desorption with 92% desorption at pH 8 (Wolska et al., 2020), and photonic/magnetic caffeine MIPs that enabled 62.5% release under 365 nm irradiation and 93.6% rebinding under visible light in real water and beverage samples (Xu et al., 2013). Magnetic, hollow, mesoporous, restricted-access, and core–shell formats further improved handling, site accessibility, separation speed, or matrix compatibility (Bhogal et al., 2021a; de Oliveira et al., 2019; Miao et al., 2015; Pajewska-Szmyt et al., 2020; Pantanit et al., 2025). MIP-based electrical/electrochemical sensors for paracetamol and moxifloxacin add architecture-focused evidence of selective, reversible, and reproducible signal transduction under common interferent challenge, although these reports are best interpreted as interferent-challenge sensing examples rather than real-sample validation where no complex sample application was reported (Shafique et al., 2024; Memoona et al., 2024). These examples make the concept of programmability concrete: imprinted soft materials can be designed not only to bind a target, but also to retrieve it magnetically, release it under a stimulus, reduce matrix interference, accelerate handling, transduce recognition events, or preserve recognition in realistic media.

A balanced interpretation also requires benchmarking and caution. Many imprinted systems outperformed non-imprinted polymers, activated carbon, XAD resin, octadecylsilane columns, Oasis hydrophilic–lipophilic-balanced sorbents, or liquid–liquid extraction in specific matrix-driven tasks (Andrade et al., 2023; Buhani et al., 2010; Herrero-Hernández et al., 2013; Krupadam et al., 2012; Li et al., 2011; Önnby et al., 2012; Tonucci et al., 2020; Venkatesh et al., 2014; Zhao et al., 2011). However, the reviewed evidence does not support treating imprinting as universally superior. UiO-66-NH₂ outperformed MIPs for artemisinin/artesunate separation, indicating that non-imprinted porous materials may be preferable for selected chemical separations (Meng et al., 2022). Ecological testing further showed that some MIPs produced negative biological effects in aquatic organisms, emphasizing that environmental compatibility must be evaluated alongside extraction performance (Guc et al., 2021). Recent electrochemical sensing work further reinforces the need for appropriate benchmarking: a non-imprinted Bi₂O₃@porous-carbon electrode achieved Hg²⁺ recovery of 99.12% ± 3.45% in canal water, retained approximately 89.6% of its signal over 30 days, and reduced material costs by more than 90% relative to noble-metal or multistep systems (Mushtaq et al., 2026). DFT-guided, molecular-modeling, virtual-imprinting, and rational-design approaches contribute to this more rigorous field direction by offering predictive or optimization frameworks that reduce trial-and-error development while improving target recognition in selected IIP and MIP systems (Lai et al., 2019; Marc et al., 2018; Rahangdale et al., 2016; Silva et al., 2018; Wang et al., 2025). Overall, the reviewed real-sample evidence supports programmable imprinted soft materials as versatile platforms for selective function in complex media, provided that claims of utility are grounded in matrix-relevant validation, appropriate comparison, regeneration performance, and clear evidence of selective function.

Limitations and future perspectives

Despite substantial progress, the field remains constrained by uneven validation standards. Many studies demonstrate selective binding or improved extraction under controlled conditions, but fewer establish performance across diverse real samples, variable matrix compositions, long-term operation, and independent benchmarking. Because selectivity can be overstated when comparisons are limited to non-imprinted analogues alone, future studies should incorporate practical competitors such as activated carbon, ion-exchange media, commercial SPE sorbents, metal–organic frameworks, porous carbons, low-cost electrodes, and established chromatographic workflows where relevant (Andrade et al., 2023; Herrero-Hernández et al., 2013; Meng et al., 2022; Tonucci et al., 2020; Venkatesh et al., 2014; Zhao et al., 2011; Mushtaq et al., 2026). This is particularly important for deployment-oriented claims, where the decisive question is not whether imprinting works, but whether it offers a measurable advantage in selectivity, cost, robustness, sustainability, or operational simplicity.

A second limitation is the inconsistent relationship between material characterization and functional interpretation. SEM, FTIR, energy-dispersive X-ray spectroscopy, XRD, BET analysis, thermogravimetric analysis, vibrating-sample magnetometry, X-ray photoelectron spectroscopy, transmission electron microscopy, and related techniques are widely used to verify morphology, functional groups, porosity, thermal behavior, magnetic response, and hybrid composition (Abdullah et al., 2019; Abdullah et al., 2020; Abdullah et al., 2023; Aboufazeli et al., 2014; Abril et al., 2026; Ansari and Masoum, 2018; Ciarlantini et al., 2025; Gao et al., 2024; Geng et al., 2024; Jalilian and Taheri, 2018; Lang et al., 2022; Li et al., 2021; Liu et al., 2013; Mamman et al., 2021; Panjali et al., 2015; Pantanit et al., 2025; Rahman et al., 2017; Tonucci et al., 2020; Wang et al., 2016; Wei et al., 2018; Yang et al., 2012; Yang et al., 2011). However, structural confirmation does not by itself prove recognition specificity. Future work should more tightly connect characterization to binding-site accessibility, site heterogeneity, diffusion behavior, template removal, nonspecific adsorption, and competitive recognition. Coupling spectroscopic and morphological evidence with kinetic models, isotherm analysis, thermodynamics, computational simulations, and rigorous interferent panels would strengthen claims that material architecture is genuinely responsible for selective function (Abdullah et al., 2019; Abdullah et al., 2020; Abdullah et al., 2023; da Silva et al., 2017; Lai et al., 2019; Rahangdale et al., 2016; Savigni et al., 2025; Silva et al., 2018).

Scalability and manufacturability also remain underdeveloped. Many imprinted materials are synthesized through multistep procedures involving specialized supports, magnetic cores, nanomaterials, controlled polymerizations, surface grafting, or hybrid recognition components. These architectures can improve performance but may complicate reproducibility, batch-to-batch consistency, cost, and regulatory translation. Promising examples such as reusable MEPS devices, magnetic extraction platforms, monoliths, straw-supported IIPs, and device-compatible sensors show that operationally efficient formats are possible (Andrade et al., 2023; Pantanit et al., 2025; Rahman et al., 2017; Sartore et al., 2020; Wang et al., 2016; Shafique et al., 2024; Memoona et al., 2024). Future research should prioritize scalable synthesis, green solvents, reduced template leakage, low-toxicity monomers, simplified regeneration, and life-cycle-aware evaluation, especially for materials intended for water treatment, food contact, clinical analysis, or environmental release.

Environmental and biological safety require greater attention. Some imprinted materials have been evaluated primarily through analytical metrics, with limited assessment of polymer leachables, nanomaterial release, degradation products, ecological effects, or post-use disposal. Evidence that certain MIPs may produce adverse effects in aquatic organisms highlights the need to evaluate environmental compatibility alongside extraction or remediation performance (Guc et al., 2021). For biomedical and food applications, additional work is needed to assess biocompatibility, fouling resistance, sterilization tolerance, matrix-induced aging, and regulatory acceptability. The next-generation of programmable imprinted soft materials should therefore be designed not only for selectivity but also for safe operation across the full use cycle.

Future advances will likely depend on predictive, rational, and data-guided design. DFT, molecular dynamics, virtual imprinting, computational monomer screening, and rational formulation optimization have already improved the selection of monomers, ligands, solvents, and template–monomer configurations in selected systems (da Silva et al., 2017; Lai et al., 2019; Marc et al., 2018; Rahangdale et al., 2016; Savigni et al., 2025; Silva et al., 2018; Wang et al., 2025). Integrating these methods with high-throughput synthesis, machine learning, standardized datasets, and uncertainty-aware performance reporting could reduce trial-and-error development. Particularly valuable would be design rules that predict when surface imprinting, dummy-template imprinting, magnetic recovery, hydrophilic matrices, responsive networks, or transducer-integrated architectures are most appropriate for a given target–matrix–function combination.

Finally, the field should move toward application-specific performance hierarchies rather than universal rankings. Adsorption capacity, LOD, imprinting factor, selectivity coefficient, recovery, RSD, regeneration, response reversibility, and real-sample validation do not carry equal importance in every setting. A water-remediation sorbent requires capacity, regeneration, safety, and cost; a clinical extraction material requires matrix tolerance, precision, and low nonspecific binding; and a sensor requires stable transduction, interferent resistance, reversibility, and calibration robustness. Future studies should therefore report performance in ways that match the intended function. Such discipline will help programmable imprinted soft materials mature from adaptable laboratory platforms into credible technologies for selective operation in complex media.

Conclusion

Programmable imprinted soft materials have advanced from template-shaped sorbents into multifunctional recognition platforms capable of selective extraction, enrichment, remediation, purification, recovery, controlled release, and sensing in complex media. Their distinctive value lies in the deliberate coordination of target identity, recognition chemistry, material architecture, operating environment, and functional readout. Across environmental, food, biomedical, industrial, and resource-oriented applications, the most persuasive examples show that imprinting can convert molecular or ionic recognition into practical matrix-tolerant performance.

The future impact of this field will depend on rigor rather than breadth alone. Stronger benchmarking, predictive design, scalable synthesis, safety evaluation, and application-specific validation are essential to distinguish true programmable selectivity from nonspecific adsorption or unnecessary material complexity. When these criteria are met, imprinted soft materials offer a powerful route toward synthetic recognition systems that are not merely selective in principle, but functional, reusable, and deployable in the chemically demanding environments where selectivity matters most.

Generative AI statement

The author(s) declared that generative AI was used in the creation of this manuscript. The authors acknowledge the use of several non-AI and AI-assisted platforms during the preparation of this work, including Web of Science, EndNote, ChatGPT Plus, Gemini, and NotebookLM. These tools were used to support literature organization, reference management, clarity, accuracy checking, structural refinement, and figure preparation. The authors have reviewed and verified the content and take full responsibility for the accuracy, integrity, and interpretation of the information presented in this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Fabricio Machado, University of Brasília, Institute of Chemistry, Brazil

Reviewed by: Arshdeep Chopra, Lingayas University, India

Nauman Sadiq, Quaid-i-Azam University, Pakistan

Footnotes

AIBN, azobisisobutyronitrile; BET, Brunauer–Emmett–Teller, DFT, density functional theory, EGDMA, ethylene glycol dimethacrylate, FTIR, Fourier-transform infrared spectroscopy, HPLC, high-performance liquid chromatography, IIP, ion-imprinted polymer, LC-MS/MS, liquid chromatography-tandem mass spectrometry, LOD, limit of detection, MEPS, microextraction by packed sorbent, MIP, molecularly imprinted polymer, RSD, relative standard deviation, SEM, scanning electron microscopy, SPE, solid-phase extraction, SPME, solid-phase microextraction, XRD, X-ray diffraction.