Molecular imprinting-based indirect fluorescence detection strategy implemented on paper chip for non-fluorescent microcystin

The study addresses the challenge of fluorescently detecting non-fluorescent environmental toxins, focusing on microcystins (particularly MC-RR) produced by cyanobacteria that pose significant health risks. Traditional fluorescent nanosensors rely on direct interactions (quenching/enhancement, FRET, PET, IFE) between targets and fluorophores, limiting applicability to species that lack such interactions and suffering from environmental susceptibility and poor selectivity. Molecularly imprinted polymers (MIPs) can confer selectivity but still typically require direct analyte–fluorophore interaction. Existing microcystin detection methods (ELISA, LC–MS/MS, HPLC, electrochemical, SPR) are accurate but complex and costly for on-site water monitoring; many fluorescence sensors use antibodies/aptamers that can be unstable, costly, and less reproducible. The purpose of this study is to develop and implement an indirect fluorescent sensing mechanism (IFSM) on a paper-based microfluidic device using MIP-coated ZnFe2O4 nanoparticles as a quencher to modulate CdTe quantum dot fluorescence, enabling sensitive, selective, rapid detection of non-fluorescent MC-RR without sample pretreatment and to demonstrate a platform generalizable to other non-interacting targets.

The paper reviews fluorescent nanoprobes (quantum dots, nanoclusters, nanodots, polymer dots) and their sensing mechanisms based on direct target–probe interactions, highlighting issues of environmental susceptibility, selectivity, and inapplicability to non-interacting targets. It discusses molecular imprinting technology to create selective recognition sites and its integration with fluorescent probes, typically relying on direct quenching/enhancement mechanisms (static/dynamic quenching, FRET, PET, IFE). Indirect approaches such as competitive binding of fluorescent indicators on MIPs have been reported (e.g., CdSe/ZnS QD assays for triazophos). For microcystins, numerous detection methods exist (ELISA, LC–MS/MS, HPLC, electrochemical, SPR), and fluorescent nanosensors using antibodies/aptamers have shown potential but face drawbacks (denaturation, environmental sensitivity, cost, reproducibility). Evolving MIP strategies include epitope and dummy template imprinting. Paper-based microfluidic analytical devices (µPADs) offer low-cost, user-friendly platforms suitable for environmental analysis, and prior studies by the authors leveraged MIP-based fluorescence sensing on paper chips. This background motivates a new indirect MIP-based fluorescent mechanism to expand applicability to non-fluorescent analytes like MC-RR.

Overview: An indirect fluorescent sensing mechanism (IFSM) is implemented on a slidable-clip paper-based microfluidic device (µSPAD) by coupling a thin MIP film-coated ZnFe2O4 nanoparticle quencher (ZnFe2O4@MIPs) with amino-functionalized CdTe quantum dots (QDs) immobilized on paper. Binding of MC-RR into imprinted cavities blocks electron/energy transfer between ZnFe2O4 and QDs, yielding fluorescence recovery proportionate to MC-RR concentration. Key materials and synthesis: - ZnFe2O4 nanoparticles (NPs): Hydrothermal synthesis. ZnO (0.192 g), FeCl3·6H2O (0.776 g), ascorbic acid (0.424 g) dispersed in 32 mL water; add 8 mL hydrazine hydrate; stir 20 min; transfer to Teflon-lined autoclave; 180 °C, 12 h; wash and dry. - MIP coating (ZnFe2O4@MIPs): Free-radical polymerization using L-arginine (dummy fragment template of MC-RR), acrylic acid (AA, functional monomer), N,N′-methylenebisacrylamide (MBA, cross-linker), and K2S2O8 (initiator). Mix 5 mg ZnFe2O4 NPs, 50 mg Arg, 35 µL AA in 15 mL water; pre-complex 120 min under N2; add 55 mg MBA; after 15 min add 10 mg K2S2O8; stir at 40 °C under N2 in the dark ~15 h. Remove template by washing with ethanol/acetic acid (9:1 v/v), with optimized elution time 24 h; wash with water. Prepare NIP similarly without Arg. Optimal synthesis conditions: 40 °C and 700 rpm stirring. - CdTe QDs: NaHTe (from 40 mg NaBH4 and 38.3 mg Te) injected into Cd precursor (92.4 mg Cd(NO3)2·4H2O, 63 µL TGA, 75 mL water), pH 9.0; heat to 96 °C and reflux 1 h; bright green emission at 530–545 nm. - Amino-functionalized paper fluorescence substrate (paper@QDs-NH2): Activate Whatman No.1 filter paper (15×15 mm) in 1 M NaOH in Teflon-lined autoclave at 120 °C for 2 h; wash. Silanize in ethanol/water (50%) with 200 µL APTES for 8 h; wash. EDC/NHS coupling: add 6 mL of 20 mg/mL EDC to 8 mL QDs, stir 5 min; add 6 mL of 10 mg/mL NHS; incubate with four paper pieces under oscillation 8 h in dark to graft QDs (paper@QDs); rinse. Further amination by immersing paper@QDs in ethanol/water (50%) with 10 µL APTES for 5 h to yield paper@QDs-NH2. - Assembling sensing site (PQ-ZnFe2O4@MIPs): Soak paper@QDs-NH2 in ZnFe2O4@MIPs dispersion; optimized MIP concentration 31.25 mg/L; soak time 15 min; dry; cut to 5×5 mm sensing pads. Device design and fabrication: - Slidable-clip µSPAD: Two identical wax-printed paper chips (50×40 mm) with hydrophobic barriers and hydrophilic channels (sample pool 5 mm dia; 8 mm channel to 5×5 mm analysis pool). Two sensing pad regions per chip and window holes; C-type slits for sliding assembly. Print with solid-wax printer (XEROX Phaser 8560DN); melt at 150 °C for 30 s to form barriers. Affix PQ-ZnFe2O4@MIPs pads to sensing sites; interleave the two chips to assemble a dual-site device. Operation protocol: - Slide chips so sample pool of one chip aligns with window of the other; deposit 15 µL sample into sample pool. Buffer: 10 mM HEPES pH 7 (optimal vs PBS/Tris). Allow equilibration for 12 min (optimized). Slide to expose sensing site in window; measure fluorescence (excitation 396 nm; emission 450–650 nm). Total analysis time <20 min. Dual sites can be run simultaneously on opposite sides without cross-interference. Optimization and characterization: - Template/monomer/cross-linker: optimal 50 mg Arg, 35 µL AA, 55 mg MBA to balance film thickness and stability. - Elution: 24 h optimal for template removal without damaging cavities. - MIP layer quality: TEM/HRTEM show ZnFe2O4 size 5–10 nm; MIP thickness ~2–3 nm. XRD confirms spinel ZnFe2O4 and amorphous polymer at ~25°. UV–Vis shows higher absorbance for MIPs vs NIPs due to cavities; FT-IR confirms polymer functional groups. TGA indicates stability below 300 °C and ~3.5% mass loading of QDs and ZnFe2O4@MIPs on paper. - Mechanism studies: IFE, FRET, and PET quenching observed between ZnFe2O4@MIPs and QDs; binding of Arg/MC-RR in cavities reduces absorbance and blocks electrostatic/hydrogen-bond interactions, leading to fluorescence recovery. Zeta potentials: QDs-NH2 +11.1 mV; ZnFe2O4@MIPs −31.6 mV (pH 7). CdTe QDs hydrated size ~5.5 nm. Lifetime measurements show decreased lifetime upon quenching and recovery upon Arg binding for MIPs, not NIPs. Real sample analysis: - Seawater (Yellow Sea, Yantai) diluted 3× in HEPES pH 7; lake water (Yantai Sanyuanhu) tested at natural pH 7. Spiked recoveries at 1.0, 5.0, 10.0 µg/L. Additional validation in seven locations from northern Taihu Lake versus LC–MS showed consistent results without sample pretreatment.

- Developed an indirect fluorescent sensing mechanism (IFSM) using ZnFe2O4@MIPs to modulate CdTe QD fluorescence, enabling detection of non-fluorescent MC-RR by blocking electron/energy transfer when imprinted cavities are occupied. - Analytical performance: Linear response for MC-RR from 0.5–50 µg/L with regression y = 0.014 x + 1.063 (R = 0.955). Limit of detection (S/N>3): 0.43 µg/L. Equilibration time: 12 min; total analysis time <20 min; sample volume 15 µL; no sample pretreatment required. - Selectivity: Minimal fluorescence enhancement for 13 potential interferents (MC-LR, amino acids, pesticides, BPA, 2,4-D, Tris), with AP and 2,4-D causing quenching rather than recovery. Saline ions showed no interference. High anti-interference capability attributed to MIP cavity recognition (shape and functional complementarity). - Stability and reproducibility: Device-to-device tests (n=20) showed low RSDs (6.54% and 2.14% for two evaluated metrics). Temperature of water samples (4–35 °C) had no significant effect. - Real samples: Spiked recoveries in seawater and lake water ranged 92.1%–109.8% with RSDs 4.41%–6.77%. Field validation at seven Taihu Lake sites showed results consistent with LC–MS, while offering rapid, pretreatment-free on-site analysis. - Mechanistic evidence: TEM/HRTEM indicated ZnFe2O4 NP size 5–10 nm and MIP layer thickness ~2–3 nm. UV–Vis showed stronger absorbance for MIPs due to cavities; absorbance decreased upon Arg rebinding. Zeta potentials (pH 7): QDs-NH2 +11.1 mV; ZnFe2O4@MIPs −31.6 mV; ZnFe2O4 +8.86 mV. Fluorescence lifetimes decreased with MIPs quenching and partially recovered upon Arg addition, supporting dynamic quenching (FRET) and PET contributions alongside IFE. The QD conduction band (−1.6 V vs NHE) makes electron transfer to Zn/Fe sites thermodynamically favorable. - Optimization highlights: Optimal MIP synthesis at 40 °C and 700 rpm; optimal elution 24 h; best buffer 10 mM HEPES pH 7; optimal ZnFe2O4@MIPs soaking concentration 31.25 mg/L and 15 min attachment; adsorption equilibrium for detection reached in ~12 min.

The findings demonstrate a practical solution to fluorescently detect non-fluorescent analytes like MC-RR by decoupling recognition from direct fluorophore–analyte interactions. The MIP-coated ZnFe2O4 acts as a selective, controllable quencher: empty imprinted cavities enable electrostatic/hydrogen-bond mediated proximity and charge/energy transfer (IFE/FRET/PET) to quench QDs; target binding fills cavities, reduces interactions, and partially restores fluorescence for quantification. This mechanism overcomes the limitations of direct fluorescent sensing when targets lack intrinsic fluorescence or quenching capability. Implementing the IFSM on a slidable-clip µSPAD yields a portable, low-cost, and user-friendly platform capable of analyzing complex water samples without pretreatment, with trace-level sensitivity and short assay time. The strong selectivity against structurally related compounds and matrix tolerance underscores the effectiveness of dummy-fragment imprinting (arginine) for MC-RR recognition. Consistency with LC–MS in field samples validates accuracy, while the absence of biological receptors (antibodies/aptamers) enhances robustness and lowers cost. The approach is generalizable to other targets that can be imprinted to control electron transfer pathways to/from fluorescent substrates.

A universal indirect fluorescent sensing strategy was established by integrating a thin MIP film on ZnFe2O4 nanoparticles with CdTe QDs on a paper-based microfluidic chip to detect non-fluorescent microcystin-RR. The device achieved sensitive (LOD 0.43 µg/L), selective, rapid (<20 min) detection over 0.5–50 µg/L with no sample pretreatment and showed strong performance in real waters and field validation against LC–MS. The IFSM, based on blocking electron/energy transfer via MIP cavity occupancy, broadens the applicability of MIP–fluorescence sensors to targets without direct fluorophore interactions. Future work should focus on synthesizing higher-affinity MIP coatings to further enhance sensitivity and on improving anti-interference capacity to extend applicability across diverse complex matrices and analyte classes.

The IFSM requires a strong interaction between the target molecule and the imprinted cavities to effectively block electron/energy transfer; weaker binding will reduce sensitivity. Enhancing the affinity of MIP coatings is necessary for improved performance. Further improvement of the system’s anti-interference capability is also needed for broader application. Additionally, due to inner-filter effects from ZnFe2O4 absorption, fluorescence recovery cannot reach the original unquenched intensity, potentially limiting dynamic range.