A conducting polymer-based array with multiplex sensing and drug delivery capabilities for smart bandages

The study addresses the need for individualized, real-time wound monitoring and on-demand therapy, especially for chronic or slow-healing wounds susceptible to infection. Traditional assessment via visual inspection or laboratory tests is qualitative or time-consuming. Key biomarkers in wound exudate, such as pH and uric acid (UA), change dynamically with healing stage and infection status. Wound pH typically shifts from mildly acidic during healing to alkaline with prolonged healing or high bacterial burden, but pH alone cannot discern infection versus tissue damage. UA increases with wound severity (ATP metabolism) but decreases with bacterial infection (microbial catabolism). Many UA biosensors assume stable pH and may misestimate UA in real wounds with fluctuating pH. The research goal is to develop a flexible, integrated theranostic bandage that simultaneously measures pH and UA with pH compensation for accurate UA quantification and provides electrically triggered, localized antibiotic delivery to enable closed-loop wound management.

Prior work has explored smart dressings and flexible biosensors for wound care, including sensors for oxygen, pH, UA, and immune proteins, as well as various passive and active dressings. UA sensors in physiological fluids often assume stable pH, limiting accuracy in wound environments with variable pH. Flexible electrode arrays have been fabricated via screen-printing, metal deposition, and photolithography; laser-induced graphene (LIG) offers a low-cost, mask-free alternative for rapid prototyping on flexible substrates. Conducting polymers (CPs) such as PANI (protonation/deprotonation responsive), PEDOT (stable conductive doped state, catalytic matrix), and PPy (actuation and drug delivery via volume change) are attractive for sensing and delivery. Prussian Blue (PB) modified electrodes improve H2O2 reduction for enzymatic biosensors but often require multi-step fabrication; a one-pot PEDOT:PB approach can simplify construction. Prior drug delivery patches demonstrated on-demand release, but precise, localized, and wirelessly controlled antimicrobial delivery integrated with multiplex sensing remains underdeveloped. This work integrates these advances, addressing UA quantification under varying pH and coupling sensing with active antibiotic delivery.

Device platform and materials: A flexible LIG electrode array was fabricated on polyimide using a CO2 laser (10.6 μm). Conducting polymers were electropolymerized onto LIG for sensing and delivery. The multiplex patch was assembled in a 3D architecture with medical adhesives (VTI Wound film), patterned chamber openings (xurographic cutting), and a Kapton insulating layer. A flexible printed circuit board (FPCB) with analog front-ends, microcontroller, BLE module, and Li-ion battery provided signal acquisition, stimulation, and wireless connectivity. PANI-based pH sensor: A LIG working electrode was modified with electropolymerized polyaniline (PANI) as the H+ sensing layer; a solid-state Ag/AgCl pseudo-reference electrode was coated with a PVB-KCl layer to stabilize potential. PANI was electropolymerized by cyclic voltammetry from –0.2 to 1.0 V for 30 cycles in 1 M HCl with 0.1 mM aniline at 50 mV s−1. The Ag/AgCl reference was formed from Ag/AgCl ink, followed by drop-casting 9 μL of 10% PVB in methanol containing KCl powder (30% w/w) and conditioning in 3 M KCl for 24 h. PEDOT:PB UA biosensor: A one-pot PEDOT:PB composite (PEDOT:PB/LIG) was deposited by cyclic voltammetry from –0.5 to 1.2 V in 1 M HCl containing 10 mM EDOT and a 10 mM mixture of K4[Fe(CN)6] and FeCl3 for 10 cycles at 50 mV s−1, forming PB nanoparticles within the PEDOT matrix on the porous LIG. Uricase immobilization: 4 μL of uricase/BSA solution (25 mg mL−1 uricase, 100 U mL−1; 10 mg mL−1 BSA) was drop-cast and dried at 4 °C for 2 h, followed by 2 μL of 1 wt% chitosan (in 1% acetic acid) as a protective/entrapment layer. The optimal uricase loading was identified as 0.4 U per electrode. Amperometric UA sensing was performed at −0.2 V to minimize interference (e.g., ascorbic acid), after evaluating potentials between −0.2 and −0.1 V. PPy:Ciprofloxacin drug carrier: Polypyrrole (PPy) co-deposited with ciprofloxacin (Cipro) was prepared by potentiostatic polymerization at 0.7 V in 0.1 M HCl containing 2.0 mM pyrrole and 50 mM Cipro until 1.5 C charge was passed. Electrodes were rinsed and briefly soaked in 0.1 M KCl to remove loosely bound surface Cipro. Electrically triggered release was evaluated at various constant and pulsed potentials; quantification of Cipro in the medium used UV–Vis absorbance (~272–277 nm). Electronics and integration: The FPCB included two configurable low-power analog front-ends (LMP91000 for amperometry and LMP2100 for potentiometry), an STM32 microcontroller for ADC, processing, and control, BLE module (JDY-23A), a 170 mAh 3.7 V Li-ion polymer battery regulated by MIC5205 LDO to 3 V, and an 8-pin connector for the patch. Firmware enabled potentiometric pH readout, amperometric UA sensing, and time-controlled stimulation for drug release, with a user interface for wireless operation. Characterization and in-vitro testing: PANI and PVB layers were verified by FTIR; morphologies by SEM. Electrochemical characterization used Fe(CN)6 redox probes for reference electrode stability, CV for H2O2 reduction on PEDOT:PB, and amperometry for H2O2 and UA calibration. pH response was measured in Britton–Robinson buffer (pH 4–10, 0.1 M KCl). In-vitro testing on porcine skin with epidermis removed used artificial wound exudates at set pH (e.g., 7–9) and UA concentrations (0.1–0.6 mM). Simultaneous potentiometric (pH) and amperometric (UA) measurements were compared to a commercial pH meter and a UV–Vis colorimetric UA assay. Antimicrobial efficacy of released Cipro against E. coli was evaluated via disk diffusion (ZOI) using exudate collected after stimulation (0.6 V, 5 min).

- PANI pH sensor: Reversible potentiometric response over pH 4–10 with Nernstian sensitivity. Calibration EMF (mV) = −59.5 × pH + 42.8 (R² = 0.997, n = 3). Sensitivity −59.5 mV pH−1, close to the theoretical −59 mV pH−1. Long-term stability with drift <0.6 mV h−1 over 12 h at pH 5, 7, 9. Rapid, reversible response upon stepwise pH changes. - PEDOT:PB UA biosensor: Strong electrocatalysis toward H2O2; linear amperometric response for H2O2 from 0.01–0.16 mM (R² = 0.998) with sensitivity 315.9 μA mM−1 at −0.10 V. For UA, linear response up to 0.9 mM (R² = 0.999) with sensitivity 1.20 μA mM−1 at −0.2 V; dynamic detection up to ~12 mM. Selectivity against common interferents (e.g., ascorbic acid, glucose, cholesterol, creatinine) demonstrated. Sensor sensitivity varies with pH (6.0–8.0); pH-dependent correction factors were derived, enabling accurate UA quantification under fluctuating pH. - Simultaneous pH and UA sensing: On porcine skin with artificial exudate, pH readings matched those from a commercial pH meter; UA readings, when corrected using pH-dependent sensitivity factors, correlated well with colorimetric assay results, validating the pH compensation approach. - PPy:Cipro drug release: Total Cipro load 80.1 ± 4.1 μg (per 1 cm² electrode). Electrically triggered release increased with positive potential up to 0.6 V; 0.6 V selected as optimal. Negative potentials (−0.2 to −0.6 V) suppressed release; 0.8 V decreased release likely due to Cipro oxidation. Active (0.6 V) release achieved rapid delivery: cumulative release ratio ~74% at 18 min and ~80.8% at 210 min, versus passive release ~25.0% at 18 min and ~31.3% at 210 min (active/passive ≈ 2.58×). Pulsed negative potentials yielded minimal release (e.g., −0.6 V ~1.44 ± 0.5 μg). Disk diffusion assays showed larger ZOI for active-release exudate (3.28 ± 0.03 cm) versus passive (2.65 ± 0.01 cm), confirming preserved and enhanced antimicrobial efficacy. - System integration: The 3D multiplex patch with FPCB enabled wireless simultaneous sensing and on-demand stimulation in a bandage prototype. In-vitro operation demonstrated accurate, real-time pH and UA readouts and user-controlled drug release via a mobile interface.

The platform demonstrates that simultaneous monitoring of wound pH and UA, combined with pH-compensated UA quantification, can resolve confounding effects of variable wound pH on enzymatic biosensing, improving the accuracy of UA as a biomarker of wound severity versus bacterial infection. The PANI-based pH sensor provided robust, near-Nernstian behavior and long-term stability, forming the basis for compensation. The PEDOT:PB-based UA biosensor leveraged one-pot fabrication for efficient H2O2 electrocatalysis at low potentials, enabling selective UA detection in complex exudates. Integrating these sensors with a PPy-based electrically triggered Cipro release module created a theranostic system capable of targeted antimicrobial delivery synchronized with biomarker readings. The flexible, laser-patterned LIG electrodes and FPCB allowed low-cost fabrication, conformal skin contact, and wireless operation, pointing to practical remote wound management. In-vitro validation on porcine skin confirmed agreement with standard methods and demonstrated functional antimicrobial action, supporting translational potential. Together, these findings indicate a viable path toward closed-loop, personalized wound care systems that can distinguish infection from non-infectious severity markers and actuate therapy on demand.

A flexible smart bandage integrating conducting polymer-based components was developed to achieve multiplex sensing (pH and UA) and on-demand antibiotic delivery. Key contributions include: (1) a PANI all-solid-state pH sensor with Nernstian performance over pH 4–10 and high stability; (2) a one-pot PEDOT:PB UA biosensor with low-potential detection, linear response up to 0.9 mM, and pH-compensation enabling accurate UA quantification in dynamically varying pH; (3) a PPy:Cipro drug carrier enabling rapid, electrically triggered release with superior antimicrobial efficacy versus passive diffusion; and (4) integration of a 3D electrode patch with a wireless FPCB into a bandage prototype for in-vitro operation. Future directions include implementing closed-loop control linking biomarker thresholds to automated drug release, enhancing dosage precision to minimize passive diffusion contributions, long-term stability testing in real wound exudates, and clinical validation to correlate biomarker dynamics with infection status and healing trajectories.

- Validation is in-vitro (porcine skin, artificial exudate); in-vivo human studies are needed to confirm performance in real wound environments. - UA biosensor sensitivity is pH-dependent, requiring calibration and compensation; residual errors from intercept differences and enzyme variability may persist. - Drug release shows a non-negligible passive component; precise dosing across the dressing’s lifecycle requires further optimization to minimize diffusion-driven release. - Negative or overly high positive potentials can suppress release or degrade Cipro; stimulation windows must be carefully controlled. - Current prototype relies on user-initiated stimulation rather than a fully closed-loop control algorithm; integration of decision logic and safety interlocks remains future work. - Long-term stability and biofouling of CP-modified LIG electrodes and enzyme layers in complex exudate over extended wear were not fully characterized.