Novel Conductive and Redox-Active Molecularly Imprinted Polymer for Direct Quantification of Perfluorooctanoic Acid

This study developed a novel molecularly imprinted polymer (MIP) that is both conductive and redox-active for directly quantifying perfluorooctanoic acid (PFOA) electrochemically. We synthesized the monomer 3,4-ethylenedioxythiophene-2,2,6,6-tetramethylpiperidinyloxy (EDOT-TEMPO) for electropolymerization on a glassy carbon electrode using PFOA as a template, which was abbreviated as PEDOT-TEMPO-MIP. The redox-active MIP eliminated the need for external redox probes. When exposed to PFOA, both anodic and cathodic peaks of MIP showed a decreased current density. This observation can be explained by the formation of a charge-assisted hydrogen bond between the anionic PFOA and MIP’s redox-active moieties (TEMPO) that hinder the conversion between the oxidized and reduced forms of TEMPO. The extent of the current density decrease showed excellent linearity with PFOA concentrations, with a method detection limit of 0.28 ng·L–1. PEDOT-TEMPO-MIP also exhibited high selectivity toward PFOA against other per- and polyfluoroalkyl substances (PFAS) at environmentally relevant concentrations. Our results suggest electropolymerization of MIPs was highly reproducible, with a relative standard deviation of 5.1% among three separate MIP electrodes. PEDOT-TEMPO-MIP can also be repeatedly used with good stability and reproducibility for PFOA detection. This study provides an innovative platform for rapid PFAS quantification using redox-active MIPs, laying the groundwork for developing compact PFAS sensors.


■ INTRODUCTION
−3 Some PFAS are highly recalcitrant, bioaccumulative, and toxic. 4,5The U.S. Environmental Protection Agency has set maximum contaminant levels (MCLs) for six PFAS in drinking water, with the MCLs for perfluorooctanesulfonic acid (PFOS) and perfluorooctanoic acid (PFOA) at 4 ng•L −1 each as of April 2024. 6Reports suggest that PFAS are present in the drinking water of 200 million Americans. 7,8There is consensus that technologies for rapid PFAS detection at sub ng•L −1 concentrations are critical but lacking.
−16 This aspect is particularly important given that many PFAS exist in mixtures and interfering ions often coexist.−23 Combining MIPs with electrochemical detection has gained immense attention, which relies on changes in electrochemical signals upon binding and removal of templates from MIP cavities.−35 However, several major obstacles exist.−33 Moreover, current studies employed non-conductive MIPs (e.g., o-phenylenediamine), [31][32][33]36 limiting electron transfer and signal transduction during measurement.Additionally, several existing conductive MIPs (e.g., polypyrrole and polyaniline) are not sufficiently stable in water, leading to decreased conductivity over time. 37,38Herein, we leveraged the high conductivity and robustness of poly-3,4-ethylenedioxythiophene (PEDOT) in water and incorporated 2,2,6,6-tetramethylpiperidinyoxy (TEMPO), a redox-active N-oxyl derivative, into PEDOT for direct quantification of PFAS at sub ng•L −1 concentrations.39−43 The detection mechanism for PFOA relies on proton blocking, hindering the conversion of TEMPOH to TEMPO + upon oxidation and reduction (Scheme 1).Specifically, we synthesized and purified the monomer EDOT-TEMPO, performed electropolymerization to obtain PEDOT-TEMPO-MIP using PFOA as a template, established a calibration curve, and assessed the selectivity, stability, and reusability of PEDOT-TEMPO-MIP towards PFOA in the range that is typical in surface water.Our findings provide an innovative platform for rapid ex-situ PFAS quantification by creating a redox-active MIP, eliminating the need for external redox probes, and laying the groundwork for compact PFAS sensor development.
■ METHODS AND MATERIALS Chemicals, Material Characterization, and Chemical Analysis.All chemicals, material characterization, and chemical analysis are provided in the Supporting Information (SI) (Texts S1, S2, and S3).
Preparation of MIPs.Details on the preparation of monomer EDOT-TEMPO, PEDOT-TEMPO-MIP, and nonmolecularly imprinted PEDOT-TEMPO (PEDOT-TEMPO-NIP) are provided in Text S4. 44,45 Electrochemical Measurement.All potentials were reported relative to the Ag/AgCl reference electrode.The cyclic voltammetry (CV) scans of glassy carbon electrode, PEDOT-TEMPO-MIP and PEDOT-TEMPO-NIP were collected at a potential range of 0.0−1.5 V with a scan rate of 20 mV•s −1 in dichloromethane (DCM) solution containing 0.1 mol•L −1 tetrabutylammonium hexafluoro phosphate (TBAPF 6 ) as a commonly used nonaqueous electrolyte.DCM was selected due to its high compatibility with PFOA and TBAPF 6 and was used for MIP electropolymerization and electrochemical signal monitoring.By contrast, the rebinding of PFAS was carried out in DI or actual water samples.The direct quantification of PFOA using PEDOT-TEMPO-MIP involves (1) MIP synthesis by electropolymerization in DCM containing TBAPF 6 , (2) removal of PFOA template with DI, (3) rebinding of PFOA with MIP in DI (or actual water sample), and (4) detection of PFOA captured by MIP in an electrochemical cell containing DCM and TBAPF 6 .Steps (1) and (4) were performed in an electrochemical cell, and steps (2) and (3) were carried out in batch reactors with water.Furthermore, the reference current density (i 0 ) was recorded in CV scans in an electrochemical cell in DCM for PEDOT-TEMPO-MIP after template (PFOA) removal.Rebinding of PFOA was conducted by exposing PEDOT-TEMPO-MIP to different PFOA concentrations (i.e., 4.14 × 10 −10 to 4.14 × 10 −4 g•L −1 ) in DI water.Afterward, PEDOT-TEMPO-MIP was transferred to an electrochemical cell for CV scans in DCM to investigate the impact of PFOA rebinding on the electrochemical signal.The baseline current density (i 0 ) of PEDOT-TEMPO-MIP was taken at an anodic peak of TEMPO after PFOA template removal, whereas i was recorded after being exposed to PFOA in DI water (also referred to as "rebinding process").Three cycles of CV scans were conducted to obtain a stable electrode response (Figure S1).The average value from the second and third scans was used for each measurement.The changes in current density (i.e., Δi = i − i 0 ) were plotted against PFOA concentrations for the calibration curve.Each PFOA concentration was measured in triplicate.

■ RESULT AND DISCUSSION
Characterization of EDOT-TEMPO monomer.The successful synthesis of the EDOT-TEMPO monomer was indicated by its orange crystal (Figure S2) and by 1 H nuclear magnetic resonance (NMR) and Fourier transform infrared (FTIR) spectroscopy (Figures S3 and S4).Text S6 explains the NMR and FTIR spectroscopy in detail.
Preparation of PEDOT-TEMPO-MIP and PEDOT-TEMPO-NIP.The CV scans were collected during the electropolymerization of PEDOT-TEMPO-MIP with PFOA (Figure S5).Consistent with the literature, 44,45 we observed a slight shift in the anodic peak of EDOT from 1.2 to 1.4 V and from 0.86 to 0.90 V for TEMPO over ten scans, which was attributed to increased resistivity between the working electrode and reference electrode following the growth of PEDOT-TEMPO-MIP film. 44,46The increased current density following electropolymerization can be explained by increased pseudocapacitance of the film by incorporating more redoxactive moieties into the MIP. 44,47The glassy carbon electrode coated with the PEDOT-TEMPO-MIP shows an apparent blue color (Figure S5). 44,45−43 The peaks at 0.30 and 0.18 V suggested anodic and cathodic peaks of the PEDOT backbone. 44,45After electropolymerization, PEDOT-TEMPO-MIP was washed with DI water instead of organic solvent to remove the PFOA template.The rationale of solvent selection was explained in Text S4. 17−19 We observed an increase of 21.1% and 18.5% in current density at anodic and cathodic peaks of PEDOT-TEMPO-MIP, respectively (Figure 1A; blue vs. red curves).When PEDOT-TEMPO-MIP was exposed to 4.14 × 10 −4 g•L −1 PFOA in DI for 30 min during rebinding, a decrease of 40% and 34% in anodic and cathodic peaks current density was observed, respectively (Figure 1A; green curve).We postulate that the observed decrease in the current density in the presence of PFOA can be attributed to the interaction between the PFOA and TEMPO moieties.Specifically, the anionic PFOA (pK a = 0.5−3.8) 48,49may form a charge-assisted hydrogen bond with TEMPOH (pK a = 5.5−6.2).As a result, the conversion from TEMPOH to TEMPO + upon oxidation is hindered, causing a decrease in anodic and cathodic peaks' current density (Scheme 1).−52 PEDOT-TEMPO-NIP was prepared under the same experimental conditions as PEDOT-TEMPO-MIP, but without PFOA.The CV scan for PEDOT-TEMPO-NIP showed CV characteristics similar to those of PEDOT-TEMPO-MIP (Figure 1B; red curve).However, the current density after removing the PFOA template (Figure 1B; blue curve) or rebinding with PFOA for 30 min (Figure 1B; green curve) exhibited negligible changes at both anodic and cathodic peaks of TEMPO, suggesting a lack of PFOA molecular imprinting cavities in PEDOT-TEMPO-NIP.
The calibration curve of PFOA was constructed by obtaining Δi from the anodic peak of TEMPO after rebinding with PFOA for 5 and 240 min, respectively, at a series of concentrations (i.e., 4.14 × 10 −10 g•L −1 to 4.14 × 10 −4 g• L −1 ).The rebinding time of 5 min was selected to establish a more realistic scenario for rapid sensing, as shown in Figure 2, Figure S6, and Table S2.The regression analysis indicated a good linear relationship at both rebinding times (5 and 240 min) (i.e., R 2 = 0.98) between Δi and corresponding PFOA concentrations (Figure 2 and Figure S6).A method detection limit (MDL) of 3.26 × 10 −9 g•L −1 for PEDOT-TEMPO-MIP was determined for 5 min exposure of PFOA using seven replicates of the lowest calibration standard (i.e., 4.14 × 10 −9 g• L −1 ) (Figure 2) following an EPA standard method (Text S5). 53,54 Our results suggest that as PFOA concentrations increase, more PFOA molecules were able to block proton transfer to TEMPO + moieties, thereby hindering conversion between TEMPO + /TEMPOH and lowering TEMPO anodic peak current density.
Reproducibility, Stability, and Selectivity of PEDOT-TEMPO-MIP.The reusability of PEDOT-TEMPO-MIP was assessed by evaluating the relative change in current density of the anodic peak of TEMPO at 0.87 V during five successive measurements of the same PFOA sample (Figure S7).The relative standard deviation (RSD) of five measurements was 6.5%.Although the reference current density (i 0 ) shifted slightly from −125 to −117 μA, possibly due to some irreversible bindings of PFOA on MIPs, the current density (i) shifted accordingly through consecutive measurements (i.e., from −116 to −108 μA).Our results suggest that PEDOT-TEMPO-MIP can be reused at least five times with good stability and reproducibility.Furthermore, the reproducibility of PEDOT-TEMPO-MIP was demonstrated by measuring 4.14 × 10 −9 g L −1 PFOA using three different electrodes.The change in current density of the anodic peak of TEMPO at 0.87 V after washing and rebinding with 4.14 × 10 −9 g L −1 PFOA was observed for each MIP electrode, and values are summarized in Table S3.Our results suggest that electropolymerization of MIPs was highly reproducible with a relative standard deviation (RSD) of 5.1%.
Lastly, we evaluated the performance of PEDOT-TEMPO-MIP in a surface water sample.The chemical composition and sampling location are provided in Table S4 and Figure  ) that are small enough to occupy MIP cavities but could not hinder H + transfer between TEMPOH and TEMPO + due to the lack of ability to form H-bond.

■ ENVIRONMENTAL SIGNIFICANCE
This work demonstrates the feasibility of synthesizing PEDOT-TEMPO-MIP that can directly quantify sub ng•L −1 PFAS concentrations electrochemically without using external redox probes.Specifically, we directly quantify PFOA with  MIP by utilizing specific interactions between PFOA and binding sites that cause signal reduction.By contrast, past published methods rely on external redox probes, which quantify PFOA indirectly and are more likely to result in a false positive.−60 It is worth noting that our calculated MDL of PEDOT-TEMPO-MIP may be further improved by applying other electrochemical techniques (e.g., differential pulse voltammetry or square wave voltammetry) to reduce background noise and enhance sensitivity. 31,47oreover, a novel PFAS sensing mechanism was elucidated for the first time.Specifically, we utilized redox-active properties of PEDOT-TEMPO-MIP, namely, the conversion between TEMPOH and TEMPO + upon oxidation and subsequent reduction, to directly quantify PFAS.The aspect of directly quantifying PFAS is of significance for environmental monitoring.−33 The ability to directly quantify PFAS helps simplify component integration for sensor platforms and enables the development of portable devices.The reversible nature of the redox-active TEMPO also helps eliminate the need for reagent replenishment and lays the groundwork for continuous monitoring, which may revolutionize the field of environmental monitoring.
Lastly, the ability to rapidly quantify PFAS is highly desirable in the field.For instance, current gold-standard LC-MS/MS requires complex sample preparation and analysis protocols, which take hours to days to obtain results. 3,10,12,59−63 By contrast, PFAS detection in our proposed platform can be accomplished within a few minutes (e.g., 5 min), which lays the groundwork for rapid detection, that is urgently needed in environmental science and engineering.Although some loss of sensitivity was observed due to interfering ions in water matrices, presumably, the calibration curve can be prepared in the same water matrices but at higher concentrations than native water, which might help correct background interference.Nonetheless, our PFAS sensor can serve as a screening tool to quickly identify water bodies impacted by PFAS pollution, allowing further PFAS characterization.Future work on the scale-up feasibility of MIP-coupled electrochemical sensors, cost analysis, and potential for real-time sensing are warranted.

■ ASSOCIATED CONTENT
* sı Supporting Information

Scheme 1 .
Scheme 1. Proposed Mechanism for Electrochemical Signal Decrease upon the Interaction between PFOA Anion and Redox-Active Moieties on PEDOT-TEMPO-MIP (i.e., TEMPO: TEMPOH/TEMPO + ) that Results in Blockage of Conversion between the Oxidized and Reduced Forms of TEMPO
S8. Specifically, surface water contains a 7.3 ± 0.4 mg•L −1 nonpurgeable organic carbon (NPOC) and 7.8 ± 0.1 mg•L −1 chloride anion.The absence of PFOA and other PFAS such as PFOS, PFBS, perfluorohexanoic acid (PFHxA), hexafluoropropylene oxide dimer acid (HFPO-DA), and 6:2-fluorotelomersulfonic acid (6:2-FTS) in DI water and collected surface water sample was confirmed by LC-MS/MS.To understand the impact of water matrices on PFOA quantification using PEDOT-TEMPO-MIP, we spiked a known concentration of PFOA (4.14 × 10 −8 g•L −1 ) into the surface water sample.The relative current density change of the anionic peak of TEMPO after rebinding was monitored following the same protocol.We observed a decrease in the current density of MIP (Figure3) by 7.8 ± 2.1% in a surface water sample, which was 18.4 ± 1.1% lower than when MIP was exposed to the same concentration of PFOA in DI water.The decrease in current density in surface water compared to DI can be attributed to the presence of interfering ions (Cl − , SO 4 2−

Figure 2 .
Figure 2. Calibration curve of the current density decrease at the anodic peak of TEMPO (y-axis) vs PFOA concentrations (ranging from 4.14 × 10 −9 to 4.14 × 10 −4 g•L −1 ; x-axis).The current density was recorded from the CV scans of PEDOT-TEMPO-MIP after being exposed to PFOA for 5 min.The error bar at each point was derived from triplicate measurements.The linear regression is y = 4.91x − 4.47 (R 2 = 0.98).