Sheep anti-1-OHPG Ab and hapten-BSA conjugate were obtained from Chemitrace Consultancy (Natland, United Kingdom). Tween-20, phosphate-buffered saline (PBS), donkey anti-sheep IgG, HRP, sulfuric acid (51% w/w), hydrogen peroxide (30 wt%), N,N-dimethylformamide (DMF), potassium ferricyanide (III), TA, and silver/silver chloride (Ag/AgCl) paste were bought from Sigma-Aldrich Chemical Co. (St. Louis, MO, USA). Sodium hydroxide, potassium hexacyanoferrate (II) trihydrate were bought from Thermo Fisher Scientific Inc. (Waltham, MA, USA). Electrically conductive composite PLA filament (CDP11705) was bought from Protoplant (Vancouver, WA, USA).

3D printing was performed on Ender-3 Pro FDM printer (Creality, Hong Kong) using PLA/carbon black (CB) filament (Protopasta, Vancouver, WA, USA) to fabricate the working electrode (WE) and counter electrode (CE). The nozzle and bed temperatures were set to 210 °C and 60 °C, respectively. The electrode was fabricated with an infill density of 100%, a printing speed of 40 mm/s, and a horizontal orientation, with the working surface parallel to the build platform (X–Y plane). The sensor substrate (2 cm × 1.5 cm) with three slots (WE, CE and RE slot) was printed using Form 3+ printer (Formlabs, Somerville, MA, USA) with Flexible 80A V1 resin (Formlabs, Somerville, MA, USA). The WE and CE were integrated into the respective slots on the sensor substrate and sealed by silicone sealant (LOCTITE SI 595 clear RTV silicone adhesive sealant) to ensure only the activated electrode surface was exposed to the detected solution. Afterwards, the reference electrode (RE) was printed into the slot of the same substrate using Ag/AgCl paste via the DIW method (customized 3D-printing system with 3 Z-axis stages, Aerotech AGS 1000) to complete the biosensor.

The 3D-printed PLA/CB WEs were immersed in the DMF solution for 5, 10 and 20 min to prepare the DMF-activated electrodes. Similarly, the electrodes were immersed in 0.5 M diluted sulfuric acid and 1 M NaOH solution for 30, 60 and 180 min to prepare H₂SO₄-activated and NaOH-activated electrodes, respectively. In addition, the electrodes were immersed into the above three solutions in sequence to prepare the DMF/NaOH/H₂SO₄-activated electrodes. For the electrochemical activation procedure, the electrodes were immersed in 0.1 M PBS solution (pH 7.2), and constant voltages of 1.5, 2, and 2.5 V were first applied to the electrodes using a CHI660E electrochemical workstation (CH Instruments, Austin, TX, USA) to determine the optimal activation potential. After identifying 2.5 V as the optimal condition, different activation times (50, 100, 150 and 200 s) were further investigated at this potential to prepare electrochemical-activated electrodes. Following activation, the electrochemical performance of the electrodes was evaluated by cyclic voltammetry (CV) in the 5 mM [Fe(CN)₆]^4−/3− solution with 0.1 M KCl as the supporting electrolyte at a scan rate of 50 mV/s.

The hapten-BSA conjugate was immobilized on the WE by the drop casting and DIW methods, respectively. A total of 4 μL of 50 ng/μL hapten-BSA conjugate was slowly dropped onto the WE surface using a pipette. By adjusting the nozzle pressure and printing time, the same volume of hapten-BSA conjugate was printed onto the WE via the DIW method. The DIW process followed a circular path, with a set pressure of 10 kPa and a printing time of 4 s. The DIW process is shown in Movie S1 of the supplementary information. After drying at room temperature, the WEs prepared by the two methods were obtained.

Supplementary video related to this article can be found at https://doi.org/10.1016/j.talanta.2026.129778.

The 3D-printed electrodes, including original electrode (without hapten-BSA conjugate immobilization), the drop casting-immobilized electrode, and the DIW-immobilized electrode, were mounted onto scanning electron microscopy (SEM) stubs using carbon adhesive tabs. A thin layer of Au (10 nm) was then sputter-coated onto the samples to ensure high-quality image acquisition and to prevent charging under the electron beam during high-voltage, high-vacuum SEM imaging. The SEM imaging was performed using the FEI Quanta 200F SEM, equipped with a field emission gun electron source and a high-resolution secondary electron detector, operating in high vacuum. To obtain high-quality images with sufficient detail, the SEM measurements were taken at an accelerating voltage of 20 kV, a working distance of 9.2 mm, and a scan speed of 30 μs, with a spot size of 3 for the electron beam.

After drying at room temperature, the hapten-BSA conjugate-immobilized WEs were washed three times with PBST buffer (0.05% TWEEN-20 in PBS). After drying, 10 μL of PBST buffer containing 1% BSA was dropped onto the WE surface for blocking at 37 °C for 1 h. Afterwards, the WEs were washed three times again with PBST buffer. The anti-1-OHPG Ab was mixed with various concentrations of 1-OHPG to prepare the detection solution. Then, 5 μL of the detection solution was added onto the WE surface and incubated at 37 °C for 2 h. Subsequently, the WEs were washed three times with PBST. After connecting the biosensor to the electrochemical workstation, 50 μL of a solution containing 1 mM thionin acetate and 50 mM hydrogen peroxide was added to fully cover all three electrodes. Finally, the electrochemical signal was recorded using an electrochemical workstation to obtain the differential pulse voltammetry (DPV) curves.

The principle of the 3D-printed flexible electrochemical immunoassay biosensor is illustrated in Fig. 1. First, hapten-BSA conjugate ink was printed onto the WE using DIW method to immobilize the hapten-BSA conjugates on the WE surface (Step 1). The sheep anti-1-OHPG Ab solution was mixed with the sample containing 1-OHPG to allow binding, and the resulting mixture was then dropped onto the WE surface. Unbound sheep anti-1-OHPG Abs subsequently bound to the immobilized hapten-BSA conjugates on the WE surface, whereas the complexes of 1-OHPG and anti-1-OHPG were unable to bind (Step 2). In other words, the 1-OHPG in the sample competed with the hapten-BSA conjugates for binding to anti-1-OHPG Abs. Afterwards, donkey anti-sheep IgG bound to the sheep anti-1-OHPG Abs, with its HRP conjugate also immobilized on the electrode surface (Step 3). HRP catalyzed the oxidation of thionine to its oxidized form, thionine (ox), in the presence of H₂O₂, generating a peak in the DPV curve [29] (Step 4). In this process, more HRP resulted in more thionine (ox), which in turn produced a higher current reduction peak in the DPV recorded by electrochemical workstation (Step 5). Therefore, the concentration of 1-OHPG in the samples was inversely proportional to the peak current observed in the DPV curve.

The 3D-printed flexible electrochemical biosensor was fabricated by a strategy employing multiple 3D printing techniques (Fig. 2), wherein each printing technique was selected to meet the corresponding functional requirements of the biosensor components.

To achieve the desired function of the 1-OHPG electrochemical biosensor, the WE and CE of the biosensor should possess high electrical conductivity, durability and mechanical strength, while remaining cost-effective [53,54]. To meet these requirements, we selected PLA/CB filament and employed FDM method to fabricate the WE (reaction area: 2 mm) and CE, as shown in Fig. 2. The primary advantage of this approach is that FDM enables the use of conductive polymer composites, where PLA ensures mechanical stability and ease of printing, while CB enhances electrical conductivity, producing electrodes with electrochemical performance for sensing applications [53,55–57]. In addition, the surface of PLA-based electrodes can be chemically or electrochemically modified and activated to increase electroactive surface area, improving sensing performance [58,59].

For the sensor platform, a flexible and high-resolution substrate is required to integrate the electrodes effectively. Therefore, SLA printing with a flexible resin was employed to fabricate the miniature substrate (2 cm × 1.5 cm), as shown in Fig. 2. Through precise printing resolution, slots were printed on the substrate to match the size of the electrodes for assembling the WE and CE. Besides, the smooth surface also facilitates electrode attachment ensuring improved sealing and adhesion. Finally, the flexibility of resin provides mechanical compliance, enabling device bending without loss of structural integrity [60], and paving the way for future wearable sensor applications.

The reference electrode requires high stability, a well-defined potential, reproducibility, and compatibility with the platform. For this, we used DIW printing to directly deposit Ag/AgCl conductive paste, a well-established reference material with a stable potential in an aqueous environment [61], via our customized 3D-printing system (AGS 1000, Aerotech) with 3 Z-axis stages, as shown in Fig. 2. The dispensing tip loaded with Ag/AgCl conductive paste is marked by a blue dotted box, with an enlarged view shown in the top-right blue box in Fig. 2. DIW enables precise deposition of Ag/AgCl paste into defined geometries (shown in the red dotted box in Fig. 2), producing reproducible reference electrodes [62]. Unlike other methods, DIW with Ag/AgCl paste avoids high-temperature curing [42], preserving the integrity of flexible resin substrate and FDM-printed WE and CE.

By combining FDM for conductive electrodes, SLA for flexible structural support, and DIW for reproducible reference electrodes, this hybrid approach achieved the fabrication of a fully integrated electrochemical sensor with mechanical flexibility, stable performance and cost-efficient production. In addition to its fabrication advantages, the proposed platform also offers low material cost. The estimated material cost of each sensor is approximately $0.17, which is lower than that of conventional gold electrodes and commercial screen-printed electrodes.

Exposing the active carbon-based materials on the electrode surface can effectively improve the electrochemical performance of 3D-printed PLA electrodes [63]. The organic solvent DMF can dissolve the PLA, thereby exposing the electrochemically active sites [56]. Selective removal of PLA by NaOH-induced saponification is also a common chemical activation method [64,65]. Moreover, acid treatment can introduce oxygenated groups, including carboxyl and hydroxyl groups, to improve the hydrophilicity of electrode surface [66,67]. Kalinke et al. reported that activation with H₂SO₄ generated more oxygenated groups on the electrode surface than with HNO₃ activation due to the stronger oxidizing capacity of H₂SO₄ [63]. Therefore, DMF, NaOH and H₂SO₄ were used for chemical activation to enhance the electrochemical activity of electrode surface in this study. The redox couple [Fe (CN)₆]^4−/3− was used as a probe to assess the electrochemical behavior and surface activation of electrodes, as it is highly sensitive to changes on carbon-based electrodes [54]. As shown in Fig. 3, the CV responses of 3D-printed electrodes treated with different activation methods were evaluated with respect to the [Fe(CN)₆]^4−/3− redox probe. The original electrode exhibited poor electrochemical activity because the carbon active sites were blocked by PLA matrix, resulting in the absence of redox peak in the CV curve (black line in Fig. 3). The effects of different immersion times for DMF, NaOH and H₂SO₄ are shown in Fig. 3A, B and C, respectively. A smaller peak-to-peak separation between oxidation and reduction peaks indicates improved electrochemical performance [68]. For DMF treatment, the 10 min-treatment yielded the best performance (blue line in Fig. 3A). Longer treatment times resulted in diminished performance due to excessive PLA removal, which degraded the electrode structure, as observed in the 20-min treatment. For NaOH, the optimal activation time was 60 min (blue line in Fig. 3B), among the treatment durations of 30, 60 and 180 min H₂SO₄ activation showed a relatively modest effect, with a similar peak-to-peak separation observed for 60 min (blue line in Figs. 3C) and 180 min (green line in Fig. 3C) treatments, although the latter produced a higher current density. As shown in Fig. S1, the electrodes subjected to the combined treatment with DMF, NaOH and H₂SO₄ exhibited a smaller separation between peaks and higher current response compared with electrodes treated with a single treatment. These results indicate improved electron-transfer kinetics and enhanced electrochemical activity after the combined activation process. Based on these findings, the chemical activation times were selected as 10 min for DMF, 60 min for NaOH, and 180 min for H₂SO₄ in subsequent experiments.

Electrochemical activation parameters were further optimized (Fig. 3D and E), as previous studies have demonstrated that electrochemical activation plays a crucial role in enhancing electrochemical performance [54,69]. First, different activation potentials were evaluated, and 2.5 V (green line in Fig. 3D) was identified as the optimal potential, as it produced the smallest peak-to-peak separation in the CV curves, indicating improved electron transfer kinetics. Subsequently, the activation time was optimized under this potential (2.5 V). As shown in Fig. 3E, an activation time of 100 s (blue line) provided the best electrochemical performance. Electrochemical activation was then carried out at 2.5 V for 100 s in subsequent experiments. After all activation steps, CV measurement was performed in a 5 mM [Fe(CN)₆]^4−/3− solution with 0.1 M KCl as the supporting electrolyte at a scan rate of 50 mV/s, and the resulting curve is presented in Fig. 3F. Compared to the original electrode (black line), the activated electrode (red line) exhibits well-defined oxidation and reduction peaks with a small peak-to-peak separation of approximately 100 mV, indicating enhanced electrochemical performance. Previous studies have reported electrochemical activation at lower potentials in alkaline or acidic media, which may offer greener alternatives to solvent-assisted activation [65,70]. In this work, the selected activation protocol was adopted because it provided reliable and reproducible electrochemical performance for the printed PLA/CB electrodes. Further optimization toward greener activation conditions will be explored in future work.

The drop-casting method is widely used for surface modification of the WE in electrochemical biosensors [51,52]. Although drop-casting is simple to operate, its drawbacks, such as the coffee ring effect (CRE), low reproducibility and uneven distribution, can negatively impact biosensor performance [71]. The coffee ring effect refers to the phenomenon in which a ring-like stain forms after a droplet containing non-volatile particles dries. As the liquid evaporates, the edge of the droplet stays pinned, causing a flow that carries the particles to the outer edge, leaving a concentrated ring [72]. Obviously, CRE leads to an uneven distribution of the deposited materials on the WE surface. To address this limitation, we used DIW technique to print the hapten-BSA conjugate on the 3D-printed WE of the biosensor. Fig. 4 compares the DIW and drop-casting methods for preparing the hapten-BSA conjugate-immobilized electrodes and illustrates the process of both immobilization techniques. For the drop-casting method, 4 μL of hapten-BSA conjugate was manually dropped onto the WE surface using a pipette. However, the limited electrode surface area and the trace amount of addition hinder the uniform distribution of hapten-BSA conjugate, reducing the reproducibility of electrodes, thereby affecting the sensitivity and accuracy of detection. The DIW method is shown in Movie S1 of the supplementary information. The pump pressure was set to 10 kPa and the print time to 4 s for the DIW printer, allowing precise printing of 4 μL of hapten-BSA conjugate ink onto the WE surface. As shown in Fig. 4A, the starting position of DIW is marked with a red dot, and the DIW path is a circle, unlike the drop-casting method, which deposits only at the center of the electrodes.

To compare the effects of the two methods for immobilizing the hapten-BSA conjugates, SEM was used to observe the electrode surface. The original WE surface after multiple activations is shown in Fig. 4B. Compared to the original surface, the hapten-BSA conjugates were immobilized on the surface of WE, as shown in Fig. 4B and C. The distribution of hapten-BSA conjugates on the WE surface prepared by DIW method is more uniform than that produced by the drop-casting method. In addition, the aggregation of hapten-BSA conjugates can be observed on the surface of the drop-casting WE (red box in Fig. 4D). The uniform immobilization of hapten-BSA conjugates is mainly due to DIW’s ability to dispense hapten-BSA conjugate ink from a nozzle in the millimeter-to micrometer-scale range and deposit it with a high level of spatial control and precision [32]. Meanwhile, the nozzle moves in a circular trajectory to evenly print the ink on the WE surface, ensuring the even distribution of hapten-BSA conjugate (Fig. 4A and Movie S1). By setting the printing speed and pressure, the dispensed volume of ink can be accurately controlled, offering a higher accuracy than the manual drop-casting method and thereby improving the reproducibility of electrode fabrication. In summary, the DIW method significantly avoids the uneven distribution caused by CRE in drop-casting, which improves the distribution of immobilized hapten-BSA conjugate on the WE surface, and thereby will enhance the detection performance of the biosensor.

Six crucial conditions influencing the immunoassay performance were optimized: the amount of hapten-BSA conjugates, dilution of sheep anti-1-OHPG Abs, dilution of donkey anti-sheep IgG, incubation time, buffer pH and test duration, as shown in Fig. 5. The DPV peak current was employed to represent the signal strength generated by the 1-OHPG biosensor. As illustrated in Fig. 5A, the peak current significantly increased with the amount of hapten-BSA conjugates immobilized on the WE surface, rising from 80 ng to 200 ng. Beyond 200 ng, the current plateaued, indicating saturation of the binding sites. For the sheep anti-1-OHPG Ab (Fig. 5B), the peak current increased with dilution factor from 3000 to 5000, with no statistically significant variation observed at higher dilutions. Similarly, the peak current intensity increased with the dilution factor of donkey anti-sheep IgG from 200 to 1000, with the highest current observed at a dilution factor of 1000, beyond which further dilution resulted in no appreciable change, as shown in Fig. 5C. In Fig. 5D, the incubation time was extended from 30 to 180 min, with the peak current reaching a maximum at 120 min. No further enhancement was observed beyond 120 min, suggesting equilibrium binding had been achieved. The effect of buffer pH on detection performance is displayed in Fig. 5E, where the optimal current response was obtained at pH 7. Deviations from this pH, either more acidic or more basic, led to diminished peak current intensity. Finally, the optimal test time was determined to be 2 min as shown in Fig. 5F. The above optimal conditions were applied to the subsequent 1-OHPG biosensing process.

The analytical performance of the 3D-printed flexible electrochemical 1-OHPG biosensor is shown in Fig. 6. It was evaluated by measuring samples containing varying concentrations of 1-OHPG. As shown in Fig. 6A, the DPV curves corresponding to different 1-OHPG concentrations ranging from 0.005 to 500 ng/mL were recorded. The peak currents were observed within the potential window of −0.25 V to −0.2 V. The peak current decreases with increasing 1-OHPG concentrations, consistent with the proposed detection mechanism. As illustrated in Fig. 6B, a linear relationship was established between the peak current and the logarithm of the 1-OHPG concentration. The biosensor exhibits a linear detection range from 0.01 ng/mL to 200 ng/mL, described by the regression equation I = −6.94 log (Conc. of 1-OHPG) +21.6, with a coefficient of determination (R²) value of 0.983. Based on the signal-to-noise ratio of 3 (S/N = 3), the limit of detection (LOD) is calculated to be 0.012 ng/mL. The enhanced sensitivity can be attributed to the improved electrochemical activity of the 3D-printed PLA/CB working electrode after chemical and electrochemical activation, which increases the electroactive surface area and facilitates electron transfer. In addition, the DIW printing of hapten-BSA enables uniform immobilization on the electrode surface, improving the efficiency of antigen–antibody interactions and enhancing electrochemical signal response. To verify the stability of the 1-OHPG biosensors, the biosensors were stored at 4 °C for 1, 3, 5, 7, and 14 days. The current responses are shown in Fig. S2. After two weeks of storage, a signal retention of 91.26% was observed, demonstrating good storage stability and stable detection performance over the two-week storage. To investigate the detection performance of 1-OHPG biosensor, 1-OHPG-spiked human urine samples were analyzed to calculate recovery. The results of the spiked recovery tests are presented in Table 1. Five concentrations of 1-OHPG (100, 50, 10, 1, and 0.1 ng/mL) were spiked into human urine samples. The recovery rate ranged from 95% to 105%, with the relative standard deviation (RSD) being below 10%. These results demonstrate the feasibility and analytical reliability of the 3D-printed 1-OHPG biosensor for quantitative detection in complex biological matrices.

The detection range and LOD of various 1-OHPG detection approaches are summarized in Table 2. As shown, conventional approaches such as LC-MS/MS, GC/MS, and HPLC are widely used for 1-OHPG quantification [20,73,74]. However, these techniques often require complex sample pretreatment steps, such as elution, filtration and extraction which limit their suitability for rapid and cost-effective analysis [14,75]. In contrast, the biosensor developed in this study provides a simplified detection procedure with a significantly lower cost. It demonstrates a wide linear detection range that exceeds that of traditional methods, along with a low LOD and a stable recovery comparable to traditional analytical techniques. Lateral flow immunoassay has also been reported for urinary 1-hydroxypyrene glucuronide detection, with a detection range of 1–10 μg/L using smartphone-based signal analysis [76]. However, smartphone-based image analysis can be susceptible to ambient light and environmental conditions. In comparison, the proposed electrochemical biosensor provides quantitative detection over a wider range (0.01–200 ng/mL), while maintaining portability for field applications.