Autonomous nanocomposite electrochemical sensing of antibiotics across aquatic ecosystems

Autonomous nanocomposite electrochemical sensing of antibiotics across aquatic ecosystems | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Autonomous nanocomposite electrochemical sensing of antibiotics across aquatic ecosystems Hong-Guang Guo, Peng Chen, Feiyun Huang, Liping Luo, Jingquan Wang, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7396857/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted You are reading this latest preprint version Abstract Antibiotics are pervasive in aquatic environments, driving antimicrobial resistance and threatening ecosystem and human health, yet monitoring still depends on laboratory-based methods that preclude real-time, large-scale surveillance. Here we report an autonomous electrochemical sensing platform based on a stabilized MXene (Ti₃C₂Tₓ) nanocomposite that enables selective, ultrasensitive, and long-term monitoring of antibiotics directly in natural waters. A controlled in situ electrochemical oxidation strategy converts Ti₃C₂Tₓ into a Ti₃C₂Tₓ-TiO₂ heterostructure while preserving electrical conductivity, producing a chemically robust and catalytically active sensing interface. When integrated with reduced graphene oxide and silver nanowires, the resulting hierarchical electrode generates multi-site molecular recognition that yields distinct electrochemical fingerprints for ciprofloxacin and sulfamethoxazole. The sensor achieves nanomolar detection limits (45 nM for ciprofloxacin and 4 nM for sulfamethoxazole), maintains > 98% signal retention over 30 days, and discriminates target antibiotics from structurally related compounds and complex matrix interferents. Machine-learning analysis of time- and voltage-resolved electrochemical features enables 96.7% classification accuracy, while density functional theory reveals fundamentally different adsorption and charge-transfer mechanisms for the two antibiotics. Field validation across lakes, rivers, aquaculture effluents, and synthetic biological fluids shows strong agreement with LC-MS/MS measurements (R 2 = 0.987). When deployed on an unmanned surface vessel, the system enables continuous, autonomous monitoring over > 10 km 2 for 72 hours. Together, these results establish electrochemical fingerprinting on stabilized MXene nanocomposites as a scalable strategy for real-time surveillance of antibiotic pollution in aquatic ecosystems. Earth and environmental sciences/Environmental sciences/Environmental chemistry/Environmental monitoring Earth and environmental sciences/Environmental sciences/Environmental impact MXene-TiO₂ nanocomposite Portable electrochemical sensor Machine learning classification Unmanned vessel monitoring Aquatic antibiotic detection Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 1. Introduction The widespread occurrence of antibiotics in aquatic environments is emerging as a major threat to ecosystem integrity and public health, driving the evolution and dissemination of antimicrobial resistance across natural and engineered water systems 1 , 2 . Antibiotic contamination occurs through pharmaceutical manufacturing discharge, agricultural runoff, and incomplete wastewater treatment, resulting in persistent environmental concentrations ranging from ng/L to µg/L that foster multidrug-resistant bacteria and threaten ecosystem stability 3 , 4 . Current regulatory frameworks mandate maximum allowable concentrations of 0.1–10 µg/L for different antibiotic classes in surface waters, yet comprehensive monitoring remains severely limited by the lack of portable, field-deployable analytical technologies 5 . Conventional methods such as liquid chromatography-tandem mass spectrometry (LC-MS/MS) provide exceptional sensitivity but require sophisticated instrumentation, specialized laboratory facilities, and extended analysis times, making them impractical for real-time field monitoring 6 . Portable electrochemical sensors offer compelling advantages, including rapid response within 5 minutes, cost-effectiveness, and miniaturization potential, making them ideal for decentralized environmental surveillance 7 , 8 , 9 . Research has focused on various nanomaterials for electrode modification, including carbon nanotubes, metal-organic frameworks, noble metal nanoparticles, and two-dimensional materials such as graphene and MXene (where M = Ti, X = C, Tₓ = surface terminations such as -OH, -O, and -F) 10, 11, 12 . Among these, MXene has attracted considerable attention due to its excellent metallic conductivity, abundant surface functional groups, and large specific surface area. To enhance electrochemical performance, researchers have explored hybridization of MXene with metal oxides, metal nanoparticles, and carbon-based materials 13 , 14 , 15 . However, MXene-based sensors face a critical challenge of oxidation instability in aqueous environments, leading to conductivity degradation that severely limits long-term operational stability and field deployment potential 8 , 16 . Several strategies have been proposed to address MXene oxidation instability, including surface functionalization, protective coating, and transformation to MXene-metal oxide composites 17 , 18 . Traditional preparation methods for MXene-TiO₂ composites, such as natural oxidation, high-temperature calcination, and hydrothermal reactions, require extended preparation cycles or extreme conditions, limiting their practical implementation 19 . Furthermore, field deployment requires sensors capable of operating across diverse water matrices with varying complexity, from simple tap water to environmental samples containing organic matter, suspended particles, and interfering ions, necessitating robust anti-interference capabilities and substrate optimization strategies 20 , 21 , 22 . Here we establish a new electrochemical materials platform in which the intrinsic oxidation instability of MXenes (Ti₃C₂Tₓ) is converted into a functional phase-engineering strategy, enabling the rapid formation of a conductive Ti₃C₂Tₓ-TiO₂ heterostructure under mild electrochemical conditions. This architecture yields a chemically stable yet electronically active surface that supports multi-analyte electrochemical fingerprinting in complex aqueous environments. To address the challenges of diverse water matrix applications, we further establish a multi-component hierarchical assembly strategy integrating multidimensional nanomaterials, which provides robust anti-interference capabilities and retains potential for bio-adaptive modification to accommodate different environmental conditions. The sensor performance was systematically characterized, and initial differential pulse voltammetry (DPV) screening across five major antibiotic classes ( Fluoroquinolones , Macrolides , Tetracyclines , Sulfonamides , and β-Lactams ) identified ciprofloxacin (CIP) and sulfamethoxazole (SMX) as exhibiting the most distinctive electrochemical signatures, achieving detection limits of 45 nmol/L and 4.0 nmol/L, respectively. By integrating operando phase-engineered MXenes, electrochemical fingerprinting, machine-learning-based pattern recognition, and molecular-level density functional theory (DFT) analysis, we demonstrate a generalizable electrochemical recognition platform capable of operating across diverse, chemically complex water matrices, from laboratory buffers to real aquatic environments and autonomous field deployment. 2. Materials and Methods 2.1 Materials Synthesis and Sensor Fabrication 2.1.1 Preparation of Ti₃C₂Tₓ nanosheets Ti₃C₂Tₓ was synthesized via a two-step approach: Ti₃AlC₂ MAX phase (where M = Ti, A = Al, X = C) precursor preparation through ball-milling and high-temperature treatment, followed by selective etching with HCl/LiF and ultrasonic delamination to obtain Ti₃C₂Tₓ nanosheets (details in Supporting Information Text S1 and Figure S1 ) . 2.1.2 Fabrication of rGO/AgNWs@Ti₃C₂Tₓ (rGAM) Electrode The rGAM electrode was constructed through hierarchical assembly methodology involving three main components. Graphene oxide (GO) was synthesized via modified Hummer’s method, while silver nanowires (AgNWs) were prepared through polyol reduction. The composite assembly was achieved by: (1) solution-phase mixing of GO with AgNWs suspension, (2) screen-printing onto electrode substrate, (3) incorporation of delaminated Ti₃C₂Tₓ nanosheets through controlled dispersion mixing, and (4) electrochemical reduction of GO to reduced graphene oxide (rGO) via cyclic voltammetry (CV) in phosphate-buffered saline (PBS) to form the final rGAM electrode (details in Supporting Information Text S2 and Figure S2 ) . 2.1.3 In-situ Electrochemical Oxidation and Electrochemical Sensor (rGAMO-ECs) Fabrication The prepared rGAM electrode underwent in-situ electrochemical oxidation treatment in a custom-designed 50 mL electrochemical reactor equipped with two platinum electrodes (8 mm × 8 mm) separated by a 20 mm distance 23 . The Ti₃C₂Tₓ component was subjected to controlled electrochemical oxidation in 50 mmol/L KCl electrolyte solution under magnetic stirring with pH adjustment and precise electrolyte resistivity control. A constant DC voltage of 6.0 V was systematically applied for various durations (20, 40, 60, 90, and 120 min) while maintaining the reaction temperature at 25–30°C. During this process, the Ti₃C₂Tₓ component in the rGAM composite was partially oxidized in situ to generate uniformly distributed TiO₂ nanoparticles on the MXene surface, yielding the rGO/AgNWs@Ti₃C₂Tₓ-TiO₂ composite (designated rGAMO). The rGAMO ink was then deposited onto screen-printed three-electrode configurations via screen printing to fabricate the rGAMO-based electrochemical sensor (rGAMO-ECs), which was subsequently integrated with a portable electrochemical workstation for smartphone-assisted data acquisition. 2.2 Structural and Morphological Characterization Comprehensive structural and morphological characterization of rGAMO-ECs composites were performed using multiple complementary techniques. Transmission electron microscopy (TEM), high-resolution TEM (HRTEM), and selected area electron diffraction (SAED) examined morphology, lattice structures, and crystallinity of individual components and composites. Energy-dispersive X-ray spectroscopy (EDS) elemental mapping analyzed spatial distribution of N, C, Ag, O, and Ti elements. Micro-infrared imaging (Micro-IR) mapped carbonyl group distribution (1735 cm⁻¹) across composite interfaces. Scanning electron microscopy (SEM) with elemental mapping characterized surface morphology and elemental distribution. X-ray photoelectron spectroscopy (XPS) determined surface composition and chemical valence states. X-ray diffraction (XRD) assessed crystallographic structure. Fourier-transform infrared spectroscopy (FTIR) identified functional groups. Confocal laser scanning microscopy (CLSM) was used to observe microstructure and interfacial characteristics. Detailed instrumentation parameters and analytical conditions are provided in Supporting Information Text S3 . 2.3 Electrochemical Measurement All electrochemical experiments were performed using a CHI760 electrochemical workstation with a three-electrode system at 25 ± 2°C. For electrode material optimization (optimal loading), operational stability, anti-interference, and reproducibility tests, rGAM was used as the working electrode, a platinum sheet served as the counter electrode, and a saturated calomel electrode (SCE) was used as the reference electrode. DPV measurements were conducted in 0.1 mol/L PBS (pH 7.6) within a potential range of 0.5–1.5 V, employing a pulse amplitude of 50 mV and a pulse width of 0.06 s 24 . Electrochemical impedance spectroscopy (EIS) was carried out in a solution containing 5 mM [Fe(CN) 6 ] 3−/4− and 0.1 mol/L KCl over a frequency range from 10⁶ to 0.1 Hz. CV was performed from − 0.6 to + 0.6 V at scan rates ranging from 20 to 240 mV/s 25 (details in Supporting Information Text S4, Figure S3 and Table S1 -Table S3) . For precise antibiotic detection and real water sample analysis, rGAMO composite was used as the working electrode, platinum sheet as counter electrode, and SCE as reference electrode. Chronoamperometry (I-T) was performed at + 0.95 V vs. SCE for CIP and + 0.83 V vs. SCE for SMX, with current responses recorded for 200 s. Linear sweep voltammetry (LSV) was conducted from 0 to + 1.4 V vs. SCE at 50 mV/s 26 . All measurements were performed in 50 mM KCl 27 . 2.4 Machine Learning (ML) Classification and Model Validation The SHapley Additive exPlanations (SHAP) framework combined with XGBoost was employed for electrochemical feature analysis and antibiotic classification. Six electrochemical parameters were extracted from chronoamperometry (I-T) and linear sweep voltammetry (LSV) measurements at concentrations of 100–1000 nM. Multi-classifier benchmarking (XGBoost, Random Forest, support vector machine (SVM)) was performed with model evaluation using accuracy, precision, recall, F1-score, and area under the receiver operating characteristic curve (AUC) metrics. t-distributed stochastic neighbor embedding (t-SNE) dimensionality reduction, SHAP analysis, and third-order polynomial regression were applied to elucidate feature contributions and non-linear dependencies (details in Supporting Information Text S5) . 2.5 Antibiotic Detection and Data Analysis Stock solutions were prepared in two concentration ranges: 1-1000 nmol/L for precision quantification in environmental samples, and 100–1000 µmol/L for broad-spectrum screening applications. Peak currents were analyzed using the Randles-Ševčík equation with verification of diffusion-controlled processes confirmed through linear Ip vs. ν¹/² relationships (R² > 0.99). Detection limits were calculated as limit of detection (LOD) = 3σ/m and limit of quantification (LOQ) = 10σ/m, where σ represents the standard deviation of blank measurements and m is the calibration curve slope (details in Supporting Information Text S6) . 2.6 Real Sample Testing and Vessel-based Offshore Monitoring Real water validation was conducted across multiple aquatic environments including tap water, natural lake water (Mingyuan Lake), flowing river water (Jiang'an River), stagnant pond water, and synthetic urine samples to assess sensor performance under diverse matrix conditions. Water samples were collected using standardized sampling protocols and filtered through 0.45 µm membranes before testing. Synthetic urine was prepared according to established protocols to simulate biological matrix complexity (details in Supporting Information Text S7) . For large-scale monitoring applications, a single-beam unmanned surface vessel equipped with rGAMO-ECs was designed and deployed for autonomous offshore surveillance. 2.7 DFT Calculations DFT calculations used the Vienna Ab initio Simulation Package (VASP) with the Perdew-Burke-Ernzerhof (PBE) functional under the generalized gradient approximation (GGA) framework, a 500 eV plane-wave cutoff, and a 3×3×1 Monkhorst-Pack k-point grid 28 . Van der Waals interactions were described using DFT-D3 dispersion correction 29 . The composite surface model comprised Ti₃C₂Tₓ MXene with TiO₂ nanoparticles interfaced with rGO layers containing AgNW segments 13 . Geometric optimization was performed until forces < 0.02 eV/Å 30 . 2.8 Statistical Analysis Statistical analysis employed multiple software platforms including Origin 2021, R software (version 4.2.2), CasaXPS (version 2.3.16), and Python-based machine learning libraries, with significance threshold set at P < 0.05. Detailed methodology for all statistical procedures, machine learning model development, feature importance analysis, and method comparison is provided in Supporting Information Text S8 . 3. Results and Discussion 3.1 Construction and Characterization of rGAMO-ECs The rGAMO-ECs platform was constructed through a hierarchical assembly strategy integrating multidimensional nanomaterials for portable antibiotic sensing at trace levels (~ 10⁻⁹ mol/L). While Ti₃C₂Tₓ, rGO, and AgNWs have been extensively explored as electrode materials, their synergistic integration through controlled electrooxidation represents a novel approach to address MXene oxidation instability. The fabrication process ( Fig. 1 a-h ) begins with Ti₃AlC₂ MAX phase synthesis ( Fig. 1 a ) , followed by selective HCl/LiF etching to generate Ti₃C₂Tₓ MXene ( Fig. 1 b ) 31 , 32 . Delamination produces atomically thin nanosheets ( Fig. 1 c ) , which are integrated with rGO and AgNWs to form the rGAM composite ( Fig. 1 d ) . The critical transformation occurs through in-situ electrooxidation, generating uniformly distributed TiO₂ nanoparticles while preserving the conductive framework ( Fig. 1 e ) . This approach avoids structural damage from traditional high-temperature treatments, preserving metallic conductivity for rapid electron transfer 33 and creating a stable core-shell structure that prevents further oxidative degradation 34 . Device fabrication employs screen printing to deposit rGAMO onto three-electrode configurations ( Fig. 1 f ) , with smartphone interfaces for real-time data acquisition ( Fig. 1 g ) . Representative I-T curves show electrochemical responses across trace concentration ranges ( Fig. 1 h ) . This hierarchical assembly simultaneously addresses the restacking tendency of 2D Ti₃C₂Tₓ nanosheets and their well-documented oxidation degradation in aqueous media by intercalating 1D AgNWs as permanent spacer channels and converting the outer Ti₃C₂Tₓ layers into a self-passivating TiO₂ shell via controlled electrochemical oxidation, achieving structural stabilization without the significant conductivity loss (> 40%) typically associated with ex situ coating or thermal annealing approaches 35 . The hierarchical architecture was characterized via HRTEM ( Fig. 1 i-p ) . The rGO component exhibited wrinkled sheet morphology with interlayer spacing of 0.365 ± 0.012 nm ( Fig. 1 i, m ) , corresponding to approximately 68% reduction degree with partial retention of oxygen-containing functional groups serving as molecular anchoring sites 36 . AgNWs demonstrated a highly crystalline face-centered cubic structure with a uniform diameter of 85 ± 15 nm ( Fig. 1 j, n ) , predominantly exposing (111) facets (d-spacing: 0.236 ± 0.009 nm, JCPDS #04-0783) that provide optimal π-π stacking interactions with quinolone rings 37 . Ti₃C₂Tₓ MXene nanosheets displayed an accordion-like layered structure with an interlayer spacing of 0.98 ± 0.05 nm ( Fig. 1 k, o ) , with intact Ti-C lattice fringes (d-spacing: 0.245 ± 0.012 nm) and sharp hexagonal SAED patterns confirming high crystallinity. The composite structure indicated successful hierarchical assembly with intimate interfacial contact ( Fig. 1 l, p ) . EDS mapping confirmed homogeneous distribution of N, C, Ag, O, and Ti throughout the matrix ( Fig. 1 q ) , with nitrogen originating from polyvinylpyrrolidone (PVP) capping agents and surface amino groups providing hydrogen bonding sites 38 . Micro-IR imaging of carbonyl group distribution (1735 cm⁻¹) revealed surface functionalization patterns ( Fig. 1 r-u ) . Pure Ti₃C₂Tₓ exhibited minimal carbonyl absorption consistent with -OH/-F terminations ( Fig. 1 r ) 39 . Introduction of rGO resulted in heterogeneous carbonyl distribution at sheet edges ( Fig. 1 s ) . The AgNWs@Ti₃C₂Tₓ composite showed enhanced carbonyl density at interfaces ( Fig. 1 t ) . The complete composite displayed the highest and most uniform carbonyl distribution ( Fig. 1 u ) , indicating synergistic creation of oxygen-containing functional groups for antibiotic binding. The uniform carbonyl distribution observed in the ternary rGAMO composite, in contrast to the edge-localized patterns in binary mixtures, indicates that the hierarchical assembly generates emergent interfacial chemical environments serving as hydrogen-bond acceptors for fluoroquinolone moieties and electron-density modulators at TiO₂ active sites for sulfonamide recognition 40 . Systematic characterization of electrochemical oxidation determined optimal parameters for TiO₂ nanoparticle generation. CLSM visualization (Figure S4) revealed progressive TiO₂ formation: sparse fluorescent points at 0–20 min formed clusters at 20–60 min, then expanded into interconnected networks at 60–90 min. The color change from metallic black to gray-black reflects Ti₃C₂Tₓ to TiO₂ transformation 41 , with preferential oxidation at high-energy sites 42 , 43 . Fluorescence stabilization after 90 min indicated oxidation equilibrium. SEM analysis (Figure S5) showed transformation from pristine structures through 10–20 nm edge nanoparticles to optimal 50–80 nm networks at 90 min, with over-oxidation at 120 min generating 200–500 nm aggregates that destroyed conductivity 44 . Elemental analysis confirmed core-shell structure formation 45 , 46 , 47 . XPS analysis (Figure S6) tracked valence evolution: C-Ti bond fluctuation reflects framework reconstruction 48 , C-O increase from 12.04% to 15.07% confirms controlled oxidation, and Ti(IV) content peaked at 56.8% at 90 min representing optimal balance. Based on comprehensive characterization, 90-min electrochemical oxidation achieves ideal balance between conductivity and catalytic activity. The well-defined oxidation optimum at 90 min reflects a self-limiting, diffusion-controlled TiO₂ growth mechanism that inherently prevents the runaway oxidation commonly encountered in chemical or thermal routes, providing a practical guideline for reproducible fabrication across different MXene compositions. 3.2 Sensing Performance and ML-Assisted Validation of rGAMO-ECs Initial DPV screening across five major antibiotic classes ( Fluoroquinolones, Macrolides, Tetracyclines, Sulfonamides, and β-Lactams ) identified six responsive antibiotics, among which CIP and SMX exhibited the most distinctive electrochemical signatures, prompting their selection for comprehensive investigation (Figure S7) . Electrochemical I-T and LSV comprehensively evaluated the trace monitoring performance of rGAMO-ECs ( Fig. 2 and Table S4) . I-T response curves revealed CIP exhibited concentration-dependent characteristics with rapid negative current changes reaching equilibrium within 3–30 s ( Fig. 2 a ) . As CIP concentration increased from 100 to 1000 nM, equilibrium negative response current decreased from − 0.045 to -0.005 µA, reflecting cumulative adsorption effects on electrode surfaces 49 , 50 . SMX I-T responses ( Fig. 2 b ) displayed similar negative trends but with distinctly different kinetics: longer equilibration times (5–95 s) and larger current changes (-0.17 to -0.02 µA, 0.15 µA total amplitude vs. CIP 0.04 µA). Extended SMX equilibration indicated complex surface interactions involving gradual binding between sulfonamide groups and TiO₂ oxygen vacancies 51 , 52 . Quantitative analysis showed CIP exhibited linear response in 100–1000 nmol/L range with 45 nmol/L detection limit (IEC = 1.028×10⁻⁴ × [CIP] − 0.09563, R² = 0.9344), while SMX demonstrated 4.0 nmol/L detection limit (IEC = 2.273×10⁻⁵ × [SMX] + 0.01687, R² = 0.9120) ( Fig. 2 c ) . The 4 nM LOD achieved for SMX is among the lowest reported for non-enzymatic electrochemical detection, approaching chromatographic sensitivity while retaining the advantages of portability and rapid response. The markedly different LODs for CIP (45 nM) and SMX reflect their distinct binding thermodynamics, with SMX benefiting from stronger electrostatic interactions between its sulfonyl group and TiO₂ oxygen vacancies 53 . LSV analysis focused on response voltage characteristics at plateau current (100 µA). CIP exhibited voltage decrease with increasing concentration ( Fig. 2 d ) , with response voltages in 1.31–1.35 V range progressively decreasing from 100 to 1000 nM, indicating catalytic-type electrochemical characteristics that facilitated electron transfer 54 , 55 . Conversely, SMX showed voltage increase with concentration ( Fig. 2 e ) , with 1.22–1.25 V response range displaying upward trends indicating increased electrochemical impedance 56 , 57 . Second-order polynomial regression revealed differential responses ( Fig. 2 f ) 58 : CIP followed P EV (V) = 1.3556–3.045×10⁻⁵ × [CIP] − 3.671×10⁻⁸ × [CIP]² (R² = 0.9924), with negative first-order coefficient indicating surface adsorption facilitating electron transfer 59 ; SMX followed P EV (V) = 1.2513 + 1.157×10⁻⁴ × [SMX] − 6.713×10⁻⁸ × [SMX]² (R² = 0.9811), with positive coefficient reflecting impedance-increasing behavior 60 . These distinct mechanisms originated from molecular structure differences: CIP possessed extended π-conjugated systems enabling rapid equilibrium and electron transfer facilitation, while SMX sulfonamide groups exhibited strong TiO₂ affinity but complex surface interactions requiring extended equilibration 61 , 62 , 63 . The opposing voltage-concentration trends for CIP and SMX enable a multi-parameter electrochemical fingerprinting approach conceptually analogous to electronic nose arrays 64 , achieved here on a single electrode through the heterogeneous surface chemistry of the ternary nanocomposite, where the extended π-conjugation of CIP facilitates catalytic-type behaviour and the lone-pair donation of SMX to TiO₂ surface states increases interfacial impedance. Signal quality analysis demonstrated exceptional analytical performance. At 100 nM concentration, signal-to-noise (S/N) ratios reached 52:1 for CIP and 38:1 for SMX ( Fig. 2 g ) , far exceeding detection (S/N = 3) and quantification (S/N = 10) thresholds 65 . Superior signal clarity resulted from rGAMO-enhanced electrocatalytic activity amplifying faradaic currents while continuous conductive networks minimized resistive losses. Baseline noise characterization revealed remarkably low current fluctuations (σ = 0.5 nA root-mean-square (RMS), Fig. 2 h), representing 4.2-fold reduction versus bare glassy carbon electrodes, attributed to rGO-AgNWs hybrid networks providing parallel conductive pathways and TiO₂ nanoparticles stabilizing interfacial double-layer capacitance 66 . Systematic parameter optimization established optimal conditions: preconcentration time studies showed that peak currents reached plateau values at 60 s ( Fig. 2 i ) , achieving 2.8-fold enhancement through spontaneous adsorption driven by π-π stacking and coordination bonding 67 ; electrolyte optimization identified 50 mM KCl in PBS (pH 7.4) as optimal ( Fig. 2 j ) , providing superior peak currents through phosphate buffering and K⁺ ionic mobility 68 ; scan rate optimization determined 10 mV/s as ideal ( Fig. 2 k ) , balancing peak resolution (> 0.95) and current intensity while minimizing non-faradaic charging current 69 . The rGAMO-ECs successfully established a multi-electrochemical fingerprint recognition system utilizing differences in current change amplitudes, voltage response trends, and reaction kinetic parameters, achieving detection limits of 45 nmol/L for CIP and 4.0 nmol/L for SMX within 100–1000 nmol/L range. To further validate these electrochemical recognition mechanisms and quantify individual parameter contributions to antibiotic discrimination, comprehensive machine learning analysis combining SHAP interpretation, dimensionality reduction, and multi-classifier validation was employed ( Fig. 3 and Figure S8) . SHAP analysis revealed distinct hierarchical patterns in antibiotic discrimination capacity across concentration ranges 70 . Low concentrations (100–400 nM) dominated feature importance, with 100 nM achieving the highest mean absolute SHAP value (0.112), followed by 200 nM (0.104) and 300 nM (0.094) ( Fig. 3 a-b ) . This concentration-dependent importance decay reflects sensor saturation kinetics, consistent with Langmuir-Freundlich adsorption models. The bidirectional SHAP distribution demonstrates contrasting electrochemical mechanisms: negative values (-0.3 to -0.1) correspond to CIP's catalytic electron transfer through π-conjugated systems, while positive values (0.1 to 0.3) indicate SMX's impedance-dominated behavior from sulfonamide surface adsorption 71 , 72 . Three-dimensional feature space analysis via t-SNE revealed excellent cluster separation between CIP, SMX, and mixture samples ( Fig. 3 c ) , validating the discriminative power of electrochemical parameters 73 . Multi-classifier benchmarking demonstrated exceptional recognition accuracy: XGBoost achieved AUC = 0.987, outperforming Random Forest (0.968) and SVM (0.941) ( Fig. 3 d and Table S5) 74 . The XGBoost confusion matrix showed near-perfect classification with only a 2% misclassification rate, where CIP achieved 98% accuracy (147/150), SMX 96.7% (145/150), and mixtures 98.7% (148/150) ( Fig. 3 e and Table S6) . Learning curve analysis confirmed model generalization without overfitting, with training and validation scores converging at approximately 95% accuracy beyond 300 training samples ( Fig. 3 f ) 75 . SHAP waterfall analysis quantified individual electrochemical parameter contributions ( Fig. 3 g ) : I-T Peak current exhibited the highest positive impact (+ 0.42), followed by LSV equilibration potential (Ep, + 0.38), while I-T Area showed negative contribution (-0.25), aligning with theoretical predictions where amperometric peak responses dominate antibiotic discrimination 76 . Feature interaction analysis revealed complex non-linear relationships ( Fig. 3 h ) , with strongest interactions between I-T Width and LSV Slope (0.33), suggesting coupled electrochemical processes. In contrast to most ML-assisted sensor studies that treat models as black boxes, our SHAP-based interpretability framework reveals that low-concentration features (100–400 nM) dominate discrimination capacity within the environmentally relevant range and identifies coupled adsorption-electron transfer processes that remain invisible to single-parameter analysis 77 . Polynomial dependency fitting of electrochemical features (Figure S8 and Table S7) validated the proposed recognition mechanisms. Third-order polynomial models achieved root-mean-square error (RMSE) values of 1.177–2.266, with I-T response voltage showing optimal fitting (RMSE = 1.177, mean absolute error (MAE) = 0.934). The non-monotonic SHAP distributions across voltage ranges confirmed distinct electrochemical fingerprints: SMX exhibited positive potential shifts (1.22–1.25 V) due to sulfonamide oxidation barriers, while CIP showed negative shifts (1.31–1.35 V) from fluoroquinolone reduction facilitation 78 , 79 . The concentration-importance hierarchy (100–300 nM optimal range) provides crucial design principles for sensor optimization, suggesting measurement precision should prioritize low-concentration detection aligned with environmental monitoring requirements (typical antibiotic levels: 50–500 nM). These machine learning results collectively validate the electrochemical recognition mechanisms of rGAMO-ECs and confirm its reliable performance for trace-level antibiotic monitoring. 3.3 Selectivity, Stability, and Reproducibility of rGAMO-ECs Anti-interference performance evaluation ( Fig. 4 a ) demonstrated that the rGAMO-ECs exhibited exceptional selective recognition capability for target antibiotics CIP and SMX. The sensor achieved response voltages of 1.33 V for CIP and 1.25 V for SMX, significantly higher than responses to interferents. Importantly, the sensor showed extremely low cross-responses to common water interferents including interfering antibiotics (ampicillin, amoxicillin, tetracycline, erythromycin, chloramphenicol), small organic molecules (glucose, urea, ascorbic acid, dopamine, serotonin), and inorganic ions (Ca²⁺, Mg²⁺, Fe³⁺, Al³⁺, Cl⁻, SO₄²⁻, NO₃⁻, PO₄³⁻, CO₃²⁻, HCO₃⁻), with response voltages maintained in the low 1.00-1.08 V range. This excellent anti-interference performance stemmed from the synergistic multi-recognition site mechanism constructed on the composite material surface 80 , 81 . The ability to discriminate CIP and SMX from their structural analogues (e.g., enrofloxacin and sulfadiazine) arises from the multi-site recognition mechanism of the ternary nanocomposite, in which rGO π-domains, AgNW facets, and TiO₂ oxygen vacancy sites collectively encode molecular geometry rather than responding to a single functional group, thereby achieving lock-and-key-like selectivity without the limitations inherent to molecularly imprinted polymers 82 . Long-term stability testing ( Fig. 4 b ) validated the outstanding durability of the rGAMO-ECs sensor. During 30 days of continuous monitoring, the sensor maintained response stability of 98.5% for CIP and 98.4% for SMX, significantly superior to rGAM-ECs (95.3% and 95.7%) and rGM-ECs (92.6% and 94.2%) sensors. The exceptional long-term stability primarily resulted from TiO₂ nanoparticle introduction, which formed stable Ti-O-C and Ti-O-Ag chemical bonding networks at composite interfaces, effectively suppressing AgNW oxidative degradation during long-term electrochemical cycling while protecting MXene substrates from structural damage 83 , 84 . The > 98% signal retention over 30 days substantially exceeds the 7–14 day operational lifetimes typical of MXene-based sensors 85 , and comparative data across electrode configurations confirm that the in situ electrochemical oxidation step, rather than the mere multi-component assembly, is the primary contributor to long-term durability through formation of a self-passivating TiO₂ barrier layer. Batch-to-batch reproducibility was evaluated across 10 independently fabricated sensor batches ( Fig. 4 c ) . Response voltages were 1.324–1.347 V for CIP with relative standard deviation (RSD) of 0.63%, and 1.239–1.268 V for SMX with RSD of 0.67%. The assessment followed the IUPAC single-laboratory validation framework and the Eurachem fitness-for-purpose principle, with the Association of Official Analytical Chemists (AOAC)/Horwitz model used to contextualize acceptable precision at trace levels. Within this framework, the inter-batch RSDs indicated robust process stability and reproducibility, supporting subsequent scale-up and practical implementation 86 , 87 . These results indicated that the established sensor preparation method possessed good process stability and reproducibility, providing a solid foundation for subsequent scale-up and practical implementation. The < 1% inter-batch RSDs are competitive with commercial screen-printed electrodes (2–5% RSD) and significantly outperform most laboratory-fabricated nanocomposite sensors (5–15% RSD) 88 , which can be attributed to the self-regulating nature of electrochemical oxidation that minimizes batch-to-batch variability in the critical surface activation step. 3.4 Field Validation of rGAMO-ECs in Diverse Water Matrices To validate the practical applicability of rGAMO-ECs in real-world environments, various actual samples representing different application scenarios were systematically evaluated using the spiked recovery method ( Fig. 5 and Figures S9) . Field testing was conducted across multiple sampling sites ( Fig. 5 a-b ) , including tap water, Mingyuan Lake, Jiang'an River, pond water, and artificial urine, covering matrix types ranging from simple to complex with varying flow conditions. Portable sensing apparatus deployment ( Figure S9a, d, g ) demonstrated the system's field adaptability across diverse aquatic environments. The complete setup included high-power AC/DC mobile power supply, membraneless air pump, filtration apparatus, portable electrochemical workstation with mobile phone interface, and reaction chamber, enabling real-time on-site detection. Considering the different flow characteristics and complexity of water environments, three electrode substrates were prepared: C-rGAMO electrodes ( Fig. 5 d and Figure S9c) for moderately flowing calm lake water, Au-rGAMO electrodes ( Fig. 5 e and Figure S9f) for flowing systems such as rivers and tap water, and Pt-rGAMO electrodes ( Fig. 5 f and Figure S9i) for nearly stagnant environments like pond water rich in organic matter (algae, etc.) and synthetic urine samples. This substrate selection approach optimizes sensor performance according to flow dynamics and matrix characteristics, where carbon substrates are suitable for moderate flow conditions, gold substrates provide enhanced performance in dynamic flowing systems, and platinum substrates exhibit superior anti-fouling capabilities in stagnant, organically complex environments 89 . The sensor demonstrated good analytical performance across all tested matrices ( Fig. 5 c ) , with recovery rates maintained within the acceptable range of 85–115%. Recovery rates for CIP detection ranged from 86.4% to 113.7%, while SMX detection achieved recovery rates of 85.1% to 109.8%. Dose-response curves maintained good linearity (R² > 0.99) across the entire concentration range (0-1000 nM). Comparative analysis of filtration membranes (Figure S9j) revealed distinct color variations reflecting differences in organic matter (algae) and sediment content across water sources. The mobile phone control interface (Figure S9k) enabled real-time monitoring and data acquisition during I-T electrochemical testing, demonstrating the system's user-friendly operation and practical deployment capabilities. This substrate optimization approach aligns with current trends in portable electrochemical sensor development and helps address matrix interference, a common challenge faced by biosensors 90 , 91 . The matrix-adaptive substrate selection strategy addresses a common limitation in sensor validation where performance is assessed in only a single matrix type, and the 85–115% recovery range achieved across all matrices meets the acceptance criteria established by US EPA Method for pharmaceutical residue analysis in environmental waters 92 . Performance analysis of different water sources indicated that sensor response was closely related to matrix complexity and substrate selection. Tap water, as the simplest matrix with relatively stable ionic composition and fewer interferents, yielded the most consistent results using C-rGAMO electrodes (CIP: 94.2%-108.6%; SMX: 91.8%-106.3%). Environmental water samples from Mingyuan Lake and Jiang'an River tested with Au-rGAMO electrodes showed moderate matrix effects while maintaining acceptable analytical precision. Pond water, containing higher concentrations of organic matter, suspended particles, and various ions, represented the most complex environmental matrix and required Pt-rGAMO electrodes to achieve acceptable recovery rates (CIP: 86.4%-113.7%; SMX: 85.1%-108.9%). Artificial urine samples showed intermediate analytical performance (CIP: 93.1%-109.4%; SMX: 90.7%-107.5%), indicating good biocompatibility and anti-protein fouling capabilities of the sensor. Field application testing ( Fig. 5 g-i ) validated the practical feasibility of this approach, with the portable workstation combined with optimized electrodes enabling on-site detection capabilities for various water bodies. By matching electrode complexity to matrix requirements, this detection method maintains analytical performance while controlling costs, meeting the needs for online monitoring of pollutants in complex environments. 3.5 Vessel-Mounted rGAMO-ECs for Long-Term Autonomous Monitoring To validate the application potential of rGAMO-ECs in large-scale water monitoring, a single-beam unmanned surface vessel equipped with rGAMO-ECs was designed and constructed ( Fig. 6 ) . The vessel-based monitoring system ( Fig. 6 a-e ) integrated solar power, Global Positioning System (GPS) positioning, fourth-generation (4G) data transmission, and flow-through detection chamber, achieving 72-hour continuous autonomous operation covering > 10 km² area. The integration of chemical-specific trace sensors with autonomous surface vessels addresses a critical gap in environmental monitoring where fixed stations offer limited spatial coverage and grab sampling fails to capture contaminant plume dynamics 93 . The successful 72-hour continuous operation confirms that the rGAMO-ECs platform meets the stability and power efficiency requirements for unattended field deployment. Field validation through parallel LC-MS/MS analysis confirmed the vessel-based system's analytical reliability (Figure S10 and Table S8) . LC-MS/MS total ion chromatogram (Figure S10a) showed complete separation of SMX (3.42 min) and CIP (4.78 min) from internal standards SMX-d4 (3.38 min) and CIP-d8 (4.73 min), with peak intensities reaching 8.24×10⁵ and 1.18×10⁶ cps, respectively. MS/MS spectra confirmed molecular structures: CIP (Figure S10b) produced characteristic fragment ions m/z 314.1 (base peak, 94.8%), 288.1 (62.3%), and 245.0 (44.7%) at collision energy 15 eV; SMX (Figure S10c) yielded m/z 156.0 (86.7%), 108.0 (71.2%), and 92.0 (54.8%) at 18 eV. Calibration curves (Figure S10d) demonstrated excellent linearity across 0.3–5000 nmol/L: CIP followed y=1847x + 232 (R²=0.998), SMX followed y=2157x + 184 (R²=0.997). Method comparison (Figure S10e) showed strong correlation between rGAMO-ECs and LC-MS/MS (y = 0.962x + 3.84, R²=0.987). Bland-Altman analysis (Figure S10f) confirmed minimal systematic bias with mean difference − 11.6 nmol/L and 95% limits of agreement ± 48.7 nmol/L. The strong LC-MS/MS correlation (R² = 0.987) and 95% limits of agreement (± 48.7 nmol/L) within the regulatory-relevant concentration range provide rigorous statistical evidence that the vessel-mounted rGAMO-ECs can serve as a reliable field surrogate for laboratory chromatographic methods 94 , supporting reduced dependence on centralized analytical infrastructure. Real aquaculture effluent validation (Table S9) demonstrated sensor performance under extreme conditions. Fish pond effluents (Site A) achieved CIP recovery of 83.6–86.4% and SMX recovery of 81.4–86.3%, despite high concentrations of interfering antibiotics (oxytetracycline 45.3 nmol/L, tetracycline 28.7 nmol/L, sulfamethazine 62.4 nmol/L, enrofloxacin 15.8 nmol/L). Shrimp pond effluents (Site B) showed similar recoveries with different interferent profiles (sulfadiazine 78.2 nmol/L, sulfathiazole 34.5 nmol/L, norfloxacin 41.3 nmol/L). Matrix effects ranged from − 10.7% to -13.7%, primarily due to high organic loads (chemical oxygen demand (COD): 186–342 mg/L, five-day biochemical oxygen demand (BOD₅): 95–168 mg/L, dissolved organic carbon (DOC): 52.8–97.3 mg/L). Mixed sample simultaneous detection validated multiplexing capability: 500 nmol/L CIP and SMX mixed spikes achieved individual recoveries of 82.4% and 80.6%, respectively, with total recovery of 81.5%. Detection limits in matrices increased compared to pure water (CIP from 45 to 68.5 nmol/L, SMX from 4.0 to 8.32 nmol/L), but remained well below regulatory limits. Despite the exceptionally challenging matrix conditions of aquaculture effluents, the maintained recoveries above 80% and successful simultaneous CIP/SMX detection validate the orthogonality of the electrochemical fingerprints for resolving individual analytes in mixed contamination scenarios typical of aquaculture discharge 95 . The vessel-based system maintained stable performance under complex water quality conditions: pH 7.2–8.4, temperature 25–28°C, dissolved oxygen 3.2–5.8 mg/L, ammonia nitrogen 12.5–35.7 mg/L. These parameters represent typical intensive aquaculture conditions, validating sensor reliability in real application scenarios. Cost-benefit analysis revealed that the vessel system with monthly sensor replacement is significantly more economical than traditional sampling and LC-MS/MS analysis, while providing real-time data (< 5 minutes) versus laboratory analysis (24–48 hours). The tolerance to wide-ranging water quality parameters without recalibration reflects the buffering capacity of the multi-component electrode architecture, and when scaled to regional vessel networks, this approach could provide the continuous, spatially resolved antibiotic surveillance data required for evidence-based antimicrobial resistance mitigation policies 96 . 3.6 Theoretical Elucidation of Antibiotic-Specific Recognition Mechanisms on rGAMO-ECs Multi-stage adsorption configuration evolution of CIP and SMX molecules on rGAMO-ECs composite surfaces revealed complete mechanistic pathways from initial contact to final stable adsorption ( Fig. 7 , Table S10-S11) . In the initial stage ( Fig. 7 a ) , CIP fluorine atoms preferentially established electrostatic interactions with cationic vacancies on Ti₃C₂Tₓ surfaces, achieving initial binding energy of -1.24 eV (F: -0.78 eV, O: -0.46 eV) 97 , 98 . SMX initial recognition occurred through sulfonyl oxygen interactions with coordinatively unsaturated titanium sites on TiO₂, yielding stronger initial binding energy of -1.67 eV (sulfonyl O: -1.21 eV, oxazole ring: -0.46 eV) 99 . During adsorption progression ( Fig. 7 b ) , molecules underwent significant spatial rearrangement: CIP rotated ~ 23° enabling oxygen atoms to establish additional interactions forming dual-point anchoring (binding energy − 1.89 eV: F -0.89 eV, O -1.00 eV) 100 , while SMX experienced ~ 31° orientation adjustment with sulfur atoms becoming active centers (binding energy − 2.15 eV: sulfonyl O -1.35 eV, S -0.80 eV). In fully optimized final configurations ( Fig. 7 c ) , CIP achieved "full-contact" mode through further ~ 15° rotation (total ~ 38°), forming multi-point anchoring networks with maximum binding energy of -2.47 eV (F: -1.02 eV, O: -1.45 eV) 100 , 101 , 102 . SMX final configuration (total rotation ~ 45°) displayed complex three-point anchoring involving sulfonyl sulfur, sulfonyl oxygen, and oxazole ring oxygen, achieving optimal binding energy of -2.83 eV (sulfonyl O: -1.47 eV, S: -0.94 eV, oxazole O: -0.42 eV). Molecular orbital analysis revealed fundamental differences in electronic interactions ( Fig. 7 d ) . CIP exhibited two binding modes: "attraction-dominated" binding involving π orbital overlap with rGO (-0.73 eV) and σ-type interactions with Ti₃C₂Tₓ/TiO₂, and "competitive equilibrium" binding with balanced attractive/repulsive orbital interactions (total covalent contributions: -1.12 eV) 103 . SMX exhibited "dual-side tight attraction" characteristics with sulfonyl lone-pair orbitals forming donor-acceptor interactions with Ti₃C₂Tₓ d orbitals and oxazole π systems mixing with TiO₂ oxygen p orbitals, resulting in enhanced covalent contributions of -1.47 eV (Ti-O bonds: -0.62 eV, AgNWs coordination: -0.61 eV) 104 . Density of states (DOS) analysis ( Fig. 7 e ) showed pristine CIP and SMX with characteristic peaks at 0.5 eV and 0.8 eV above Fermi level. Upon adsorption, CIP-surface system generated new states at -2.5 eV and − 1.2 eV below the Fermi level (EF) (integrated DOS: 3.2 states/eV), indicating strong σ-bonding, while SMX introduced broader state distribution between − 3.0 and − 1.0 eV (integrated DOS: 4.1 states/eV), confirming enhanced orbital hybridization and stronger electronic coupling. The π-dominated binding of CIP through its conjugated quinolone system creates efficient charge transfer channels that manifest as a voltage-decreasing response, whereas the lone-pair-mediated donor-acceptor binding of SMX with Ti d-orbitals introduces localized electronic states that increase interfacial impedance and produce a voltage-rising response, representing two fundamentally different transduction pathways coexisting on the same heterogeneous electrode surface 105 . Potential energy surface landscapes (PES) provided critical thermodynamic and kinetic insights. For CIP ( Fig. 7 f ) , the PES revealed primary energy well at coordinates (0, -1) Å corresponding to O-Ti interaction sites, with barrier height 1.08 eV and well depth − 2.47 eV 106 . Energy gradient along the approach pathway measured − 0.36 eV/Å with characteristic funnel radius 2.8 Å facilitating molecular orientation. Secondary local minima at (± 2, 0) Å (depth − 0.72 eV) represented alternative F-Ti binding sites. SMX PES ( Fig. 7 h ) displayed a broader energy basin with primary minimum at (0, 0) Å reaching − 2.83 eV depth, and multiple secondary minima at (± 2, ± 1) Å (-1.54 eV depth) corresponding to Ring-N and SO-O coordination sites. A gentler approach gradient (-0.24 eV/Å) and extended flat region (radius ~ 3.5 Å) around primary minimum indicated lower kinetic barriers (0.85 eV) but greater configurational flexibility. Charge density difference analysis revealed distinct electronic redistribution patterns at molecule-surface interfaces. For CIP ( Fig. 7 g ) , charge accumulation regions (Δρ = +0.36 e/ų) concentrated between F atoms and Ti sites at 2.1 Å distances, with complementary depletion zones (Δρ = -0.48 e/ų) surrounding quinolone backbone. Integrated charge transfer amounted to 0.42 electrons from CIP to surface with 2.3 Debye dipole moment change. SMX charge density mapping ( Fig. 7 i ) showed more pronounced polarization with intense accumulation at sulfonyl-Ti interfaces (Δρ = +0.24 e/ų) at 1.9 Å and significant depletion around benzene ring (Δρ = -0.52 e/ų). A net transfer of 0.58 electrons and 3.1 Debye dipole moment change confirmed stronger ionic character in SMX binding. Comprehensive interaction energy analysis demonstrated electrostatic interactions contributed more to SMX binding (-0.89 eV) versus CIP (-0.52 eV), primarily through stronger S-Ti binding (-0.54 eV) versus F-Ti interactions (-0.31 eV). Van der Waals forces favored CIP binding (-0.51 eV versus SMX − 0.31 eV), while induced dipole interactions showed greater significance for CIP (-0.32 eV) than SMX (-0.16 eV), reflecting different electronic environments and polarization characteristics 107 . These quantitative findings aligned with experimental adsorption isotherms showing SMX maximum capacity of 287.3 mg/g exceeding CIP 243.7 mg/g, validating stronger theoretical binding energy predictions. 4. Conclusions This work demonstrates that the intrinsic oxidation instability of MXenes can be harnessed as a phase-engineering strategy to create stable, catalytically active sensing interfaces. The rGAMO-ECs platform achieves nanomolar detection limits (45 nmol/L for CIP and 4.0 nmol/L for SMX), > 98% signal retention over 30 days, and 96.7% machine-learning classification accuracy, while DFT calculations reveal fundamentally different attraction-dominated (CIP) and impedance-dominated (SMX) recognition mechanisms that provide molecular-level design principles for tailoring electrode surfaces. Field validation confirms strong LC-MS/MS correlation (R² = 0.987) across diverse water matrices, and autonomous vessel deployment enables 72-hour continuous monitoring covering > 10 km², further demonstrating that nanocomposite electrochemical sensors can transition from laboratory settings to large-scale environmental surveillance. Declarations Acknowledgments The authors would like to thank the National Key Research and Development Program of China (2023YFC3210100), National Natural Science Foundation of China (42177060), and Science & Technology Department of Sichuan (2023NSFSC1949, 2023ZHCGO024-LH) for the financial support. References Loffler P, Escher B, Baduel C, Virta MP, Lai FY (2023) Antimicrobial transformation products in the aquatic environment: global occurrence, ecotoxicological risks, and potential of antibiotic resistance. Environ Sci Technol 57:9474–9494 Rodriguez-Mozaz S et al (2020) Antibiotic residues in final effluents of European wastewater treatment plants and their impact on the aquatic environment. Environ Int 140:105733 Zhang Q, Ying G, Pan C, Liu Y-S, Zhao J-L (2015) Comprehensive evaluation of antibiotics emission and fate in the river basins of China: source analysis, multimedia modeling, and linkage to bacterial resistance. Environ Sci Technol 49:6772–6782 Hanna N et al (2018) Presence of antibiotic residues in various environmental compartments of Shandong province in eastern China: its potential for resistance development and ecological and human risk. Environ Int 114:131–142 Liguori K et al (2022) Antimicrobial resistance monitoring of water environments: a framework for standardized methods and quality control. Environ Sci Technol 56:9149–9160 Dorival-García, Zafra‐Gómez C, Vílchez Navalón (2013) Simultaneous determination of 13 quinolone antibiotic derivatives in wastewater samples using solid‐phase extraction and ultra performance liquid chromatography–tandem mass spectrometry. Microchem J 106:323–333 Frigoli M et al (2024) Electrochemical Sensors for Antibiotic Detection: A Focused Review with a Brief Overview of Commercial Technologies. Sensors 24:5576 Li R et al (2020) A flexible and physically transient electrochemical sensor for real-time wireless nitric oxide monitoring. Nat Commun 11:3207 Coatsworth P et al (2024) Time-resolved chemical monitoring of whole plant roots with printed electrochemical sensors and machine learning. Sci Adv 10:6315 Wu HB, Lou XW (2017) Metal-organic frameworks and their derived materials for electrochemical energy storage and conversion: Promises and challenges. Sci Adv 3:9252 Gao C, Yu X-Y, Xu R-X, Liu J-H, Huang X-J (2012) AlOOH-reduced graphene oxide nanocomposites: one-pot hydrothermal synthesis and their enhanced electrochemical activity for heavy metal ions. ACS Appl Mater Interfaces 4:4672–4682 Wang R, Wang S, Zhang Y, Jin D, Tao X, Zhang L (2018) Graphene-coupled Ti 3 C 2 MXenes-derived TiO 2 mesostructure: promising sodium-ion capacitor anode with fast ion storage and long-term cycling. J Mater Chem A 6:1017–1027 Hantanasirisakul K, Gogotsi Y (2018) Electronic and optical properties of 2D transition metal carbides and nitrides (MXenes). Adv Mater 30:1804779 Anasori B, Lukatskaya MR, Gogotsi Y (2023) 2D metal carbides and nitrides (MXenes) for energy storage. MXenes ). Jenny Stanford Publishing Naguib M et al (2023) Two-dimensional nanocrystals produced by exfoliation of Ti 3 AlC 2 . MXenes ). Jenny Stanford Publishing Yang Z et al (2024) Nanozyme-enhanced electrochemical biosensors: mechanisms and applications. Small 20:2307815 Tian W et al (2019) Multifunctional nanocomposites with high strength and capacitance using 2D MXene and 1D nanocellulose. Adv Mater 31:1902977 Nayak P, Kurra N, Xia C, Alshareef HN (2016) Highly efficient laser scribed graphene electrodes for on-chip electrochemical sensing applications. Adv Electron Mater 2:1600185 Pazniak H et al (2020) Partially oxidized Ti3C2T x MXenes for fast and selective detection of organic vapors at part-per-million concentrations. ACS Appl Nano Mater 3:3195–3204 Gopinath KP, Madhav NV, Krishnan A, Malolan R, Rangarajan G (2020) Present applications of titanium dioxide for the photocatalytic removal of pollutants from water: A review. J Environ Manage 270:110906 Reddy YVM, Sravani B, Łuczak T, Mallikarjuna K, Madhavi G (2021) An ultra-sensitive rifampicin electrochemical sensor based on titanium nanoparticles (TiO 2) anchored reduced graphene oxide modified glassy carbon electrode. Colloids Surf Physicochem Eng Aspects 608:125533 Rhouati A, Berkani M, Vasseghian Y, Golzadeh N (2022) MXene-based electrochemical sensors for detection of environmental pollutants: A comprehensive review. Chemosphere 291:132921 Kovalenko I et al (2011) A major constituent of brown algae for use in high-capacity Li-ion batteries. Science 334:75–79 Sharma N, Selvam SP, Yun K (2020) Electrochemical detection of amikacin sulphate using reduced graphene oxide and silver nanoparticles nanocomposite. Appl Surf Sci 512:145742 Li Y et al (2020) A highly sensitive and selective molecularly imprinted electrochemical sensor modified with TiO 2 -reduced graphene oxide nanocomposite for determination of podophyllotoxin in real samples. J Electroanal Chem 873:114439 Luo J et al (2017) Pillared structure design of MXene with ultralarge interlayer spacing for high-performance lithium-ion capacitors. ACS Nano 11:2459–2469 Turlybekuly A, Shynybekov Y, Soltabayev B, Yergaliuly G, Mentbayeva A (2024) The cross-sensitivity of chemiresistive gas sensors: Nature, methods, and peculiarities: A systematic review. ACS Sens 9:6358–6371 Kresse G, Furthmüller J (1996) Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Phys Rev B 54:11169 Grimme S, Antony J, Ehrlich S, Krieg H (2010) A consistent and accurate ab initio parametrization of density functional dispersion correction (DFT-D) for the 94 elements H-Pu. J Chem Phys 132 Henkelman G, Uberuaga BP, Jónsson H (2000) A climbing image nudged elastic band method for finding saddle points and minimum energy paths. J Chem Phys 113:9901–9904 Chen J et al (2020) Recent progress and advances in the environmental applications of MXene related materials. Nanoscale 12:3574–3592 Kim Y-J et al (2021) Etching mechanism of monoatomic aluminum layers during MXene synthesis. Chem Mater 33:6346–6355 Iqbal A, Hong J, Ko TY, Koo CM (2021) Improving oxidation stability of 2D MXenes: synthesis, storage media, and conditions. Nano Convergence 8:9 Soomro RA, Zhang P, Fan B, Wei Y, Xu B (2023) Progression in the oxidation stability of MXenes. Nano Micro Lett 15:108 Natu V, Clites M, Pomerantseva E (2018) Mesoporous MXene powders synthesized by acid induced crumpling and their use as Na-ion battery anodes. Mater Res Lett 6:230–235 Dreyer DR, Park S, Bielawski CW, Ruoff RS (2010) The chemistry of graphene oxide. Chem Soc Rev 39:228–240 Lazar P et al (2013) Adsorption of small organic molecules on graphene. J Am Chem Soc 135:6372–6377 Zhang CJ et al (2017) Oxidation stability of colloidal two-dimensional titanium carbides (MXenes). Chem Mater 29:4848–4856 Hope MA et al (2016) NMR reveals the surface functionalisation of Ti 3 C 2 MXene. Phys Chem Chem Phys 18:5099–5102 Seh ZW, Kibsgaard J, Dickens C (2017) Combining theory and experiment in electrocatalysis: Insights into materials design. Science 355:eaad4998 Tran NM, Ta QTH, Noh J-S (2021) Unusual synthesis of safflower-shaped TiO 2 /Ti 3 C 2 heterostructures initiated from two-dimensional Ti 3 C 2 MXene. Appl Surf Sci 538:148023 He S, Shang C, Lu H, Xu Z, Zhao H (2025) Experimental and numerical study of TiO 2 nanoparticle evolution in a diffusion flame reactor. Combust Flame 273:113965 Lu Y et al (2018) Self-hydrogenated shell promoting photocatalytic H 2 evolution on anatase TiO 2 . Nat Commun 9:2752 Shang H et al (2023) Surface hydrogen bond-induced oxygen vacancies of TiO 2 for two-electron molecular oxygen activation and efficient NO oxidation. Environ Sci Technol 57:20400–20409 Ghassemi H et al (2014) In situ environmental transmission electron microscopy study of oxidation of two-dimensional Ti 3 C 2 and formation of carbon-supported TiO 2 . J Mater Chem A 2:14339–14343 Noman M et al (2024) Ti3C2Tx-MXene based 2D/3D Ti 3 C 2 –TiO 2 –CuTiO 3 heterostructure for enhanced pseudocapacitive performance. Chem Eng J 499:156697 Xia Y et al (2018) Thickness-independent capacitance of vertically aligned liquid-crystalline MXenes. Nature 557:409–412 Liu N et al (2022) High-temperature stability in air of Ti 3 C 2 T x MXene-based composite with extracted bentonite. Nat Commun 13:5551 Zhou J, Huang S, He Z, Song S (2021) Enhanced activity and stability of PbO2 electrodes by modification with octadecyl phosphonic acid. J Electrochem Soc 168:116503 Do BV et al (2023) Voltammetric determination of a fluoroquinolone antibiotic based on multilayer reduced graphene oxide sensor prepared directly, promptly by electrochemically expanding graphite electrode surface. Int J Environ Anal Chem 103:8141–8157 Lim J, Yang Y, Hoffmann MR (2019) Activation of peroxymonosulfate by oxygen vacancies-enriched cobalt-doped black TiO 2 nanotubes for the removal of organic pollutants. Environ Sci Technol 53:6972–6980 Cao Y et al (2020) Fabrication of novel CuFe 2 O 4 /MXene hierarchical heterostructures for enhanced photocatalytic degradation of sulfonamides under visible light. J Hazard Mater 387:122021 Majdinasab M, Daneshi M, Marty JL (2021) Recent developments in non-enzymatic (bio) sensors for detection of pesticide residues: Focusing on antibody, aptamer and molecularly imprinted polymer. Talanta 232:122397 Shafiei H, Hassaninejad-Darzi SK (2023) Electroanalytical application of Ag@ POM@ rGO nanocomposite and ionic liquid modified carbon paste electrode for the quantification of ciprofloxacin antibiotic. J Electroanal Chem 935:117321 Santos AM, Wong A, Almeida AA, Fatibello-Filho O (2017) Simultaneous determination of paracetamol and ciprofloxacin in biological fluid samples using a glassy carbon electrode modified with graphene oxide and nickel oxide nanoparticles. Talanta 174:610–618 Zhu X et al (2022) The effect of sulfamethoxazole on nitrogen removal and electricity generation in a tidal flow constructed wetland coupled with a microbial fuel cell system: Microbial response. Chem Eng J 431:134070 Zhang S, Song H-L, Yang X-L, Li H, Wang Y-W (2018) A system composed of a biofilm electrode reactor and a microbial fuel cell-constructed wetland exhibited efficient sulfamethoxazole removal but induced sul genes. Bioresour Technol 256:224–231 Xu X, Wu S, Guo D, Niu X (2020) Construction of a recyclable oxidase-mimicking Fe3O4@ MnOx-based colorimetric sensor array for quantifying and identifying chlorophenols. Anal Chim Acta 1107:203–212 Li S, Zhang X, Huang Y (2017) Zeolitic imidazolate framework-8 derived nanoporous carbon as an effective and recyclable adsorbent for removal of ciprofloxacin antibiotics from water. J Hazard Mater 321:711–719 Wang S, Zhang J, Gharbi O, Vivier V, Gao M, Orazem ME (2021) Electrochemical impedance spectroscopy. Nat Rev Methods Primers 1:41 Zheng X et al (2024) Oxygen vacancies-promoted removal of COS via catalytic hydrolysis over CuTiO2-δ nanoflowers. Chem Eng J 492:152322 Zhu L, Santiago-Schübel B, Xiao H, Hollert H, Kueppers S (2016) Electrochemical oxidation of fluoroquinolone antibiotics: Mechanism, residual antibacterial activity and toxicity change. Water Res 102:52–62 Zeng W, Zhang H, Wu R, Liu L, Li G, Liang H (2023) Environment-friendly and efficient electrochemical degradation of sulfamethoxazole using reduced TiO 2 nanotube arrays-based Ti membrane coated with Sb-SnO 2 . J Hazard Mater 446:130642 Peveler W, Yazdani M, Rotello V (2016) Selectivity and specificity: pros and cons in sensing. ACS Sens 1:1282–1285 Maduraiveeran G, Sasidharan M, Ganesan V (2018) Electrochemical sensor and biosensor platforms based on advanced nanomaterials for biological and biomedical applications. Biosens Bioelectron 103:113–129 Wang X, Chen Y, Schmidt OG, Yan C (2016) Engineered nanomembranes for smart energy storage devices. Chem Soc Rev 45:1308–1330 Sun Y, He K, Zhang Z, Zhou A, Duan H (2015) Real-time electrochemical detection of hydrogen peroxide secretion in live cells by Pt nanoparticles decorated graphene–carbon nanotube hybrid paper electrode. Biosens Bioelectron 68:358–364 Ni Y, Yue W, Liu F, Bi W, Sun Z, Wu Y (2023) Efficient electrochemical oxidation of cephalosporin antibiotics by a highly active cerium doped PbO 2 anode: Parameters optimization, kinetics and degradation pathways. Colloids Surf Physicochem Eng Aspects 666:131318 Martin-Yerga D, Costa-Garcia A, Unwin PR (2019) Correlative voltammetric microscopy: structure–activity relationships in the microscopic electrochemical behavior of screen printed carbon electrodes. ACS Sens 4:2173–2180 Lundberg SM, Lee S-I (2017) A unified approach to interpreting model predictions. Adv Neural Inf Process Syst 30 Gayen P, Chaplin BP (2016) Selective electrochemical detection of ciprofloxacin with a porous nafion/multiwalled carbon nanotube composite film electrode. ACS Appl Mater Interfaces 8:1615–1626 Labib M, Sargent EH, Kelley SO (2016) Electrochemical methods for the analysis of clinically relevant biomolecules. Chem Rev 116:9001–9090 Maaten Lvd, Hinton G (2008) Visualizing data using t-SNE. J Mach Learn Res 9:2579–2605 Ester M, Kriegel H, Xu X, XGBoost: (2016) A scalable tree boosting system. In Proceedings of the 22Nd ACM SIGKDD International Conference on Knowledge Discovery and Data Miningvol, pg 785, Geogr Anal , (2022) Cui F, Yue Y, Zhang Y, Zhang Z, Zhou HS (2020) Advancing biosensors with machine learning. ACS Sens 5:3346–3364 Puthongkham P, Wirojsaengthong S, Suea-Ngam A (2021) Machine learning and chemometrics for electrochemical sensors: moving forward to the future of analytical chemistry. Analyst 146:6351–6364 Wei L, Chen Z, Shi L (2017) Super-multiplex vibrational imaging. Nature 544:465–470 Nguyen T-B et al (2023) Phosphoric acid-activated biochar derived from sunflower seed husk: Selective antibiotic adsorption behavior and mechanism. Bioresour Technol 371:128593 Igwegbe CA, Oba SN, Aniagor CO, Adeniyi AG, Ighalo JO (2021) Adsorption of ciprofloxacin from water: a comprehensive review. J Ind Eng Chem 93:57–77 Chen K, Liu S, Zhou Y, Zhao G (2025) Monitoring and Analysis of Total Tetracyclines in Water from Different Environmental Scenarios Using a Designed Broad-Spectrum Aptamer. Environ Sci Technol 59:5736–5746 Huang W et al (2025) Wearable Sensor for Continuous Monitoring Multiple Biofluids: Improved Performances by Conductive Metal-Organic Framework with Dual‐Redox Sites on Flexible Graphene Fiber Microelectrode. Adv Funct Mater, 2424018 BelBruno JJ (2018) Molecularly imprinted polymers. Chem Rev 119:94–119 Chen J et al (2023) Reverse oxygen spillover triggered by CO adsorption on Sn-doped Pt/TiO 2 for low-temperature CO oxidation. Nat Commun 14:3477 Bhat A, Anwer S, Bhat KS, Mohideen MIH, Liao K, Qurashi A (2021) Prospects challenges and stability of 2D MXenes for clean energy conversion and storage applications. npj 2D Mater Appl 5:61 Soomro RA, Jawaid S, Zhu Q, Abbas Z, Xu B (2020) A mini-review on MXenes as versatile substrate for advanced sensors. Chin Chem Lett 31:922–930 Borman PJ et al (2024) Ongoing analytical procedure performance verification using a risk-based approach to determine performance monitoring requirements. Anal Chem 96:966–979 Hernández F, Fabregat-Safont D, Campos-Mañas M, Quintana JB (2023) Efficient validation strategies in environmental analytical chemistry: A focus on organic micropollutants in water samples. Annu Rev Anal Chem 16:401–428 Hayat A, Marty JL (2014) Disposable screen printed electrochemical sensors: Tools for environmental monitoring. Sensors 14:10432–10453 Bunnasit S, Thamsirianunt K, Rakthabut R, Jeamjumnunja K, Prasittichai C, Siriwatcharapiboon W (2024) Sensitive portable electrochemical sensors for antibiotic chloramphenicol by tin/reduced graphene oxide-modified screen-printed carbon electrodes. ACS Appl Nano Mater 7:267–278 Silva FWL et al (2024) Disposable electrochemical sensor: Highly sensitive determination of nitrofurazone antibiotic in environmental samples and pharmaceutical formulations. Chemosphere 361:142481 Khezerloo E, Hekmat F, Zad AI, Shahrokhian S (2025) A novel electrochemical probe crafted from copper oxide Nanoball-multiwalled carbon nanotube for valacyclovir detection in environmental and medicinal matrices. Electrochim Acta, 146872 Mao K, Zhang H, Pan Y (2021) Biosensors for wastewater-based epidemiology for monitoring public health. Water Res 191:116787 Dunbabin M, Marques L (2012) Robots for environmental monitoring: Significant advancements and applications. IEEE Rob Autom Magazine 19:24–39 Thompson M, Ellison SL, Wood R (2002) Harmonized guidelines for single-laboratory validation of methods of analysis (IUPAC Technical Report). Pure Appl Chem 74, 835–855 Lulijwa R, Rupia EJ, Alfaro AC (2020) Antibiotic use in aquaculture, policies and regulation, health and environmental risks: a review of the top 15 major producers. Rev Aquacult 12:640–663 Larsson DJ, Flach C-F (2022) Antibiotic resistance in the environment. Nat Rev Microbiol 20:257–269 Näslund L-Å, Mikkelä M-H, Kokkonen E, Magnuson M (2021) Chemical bonding of termination species in 2D carbides investigated through valence band UPS/XPS of Ti 3 C 2 T x MXene. 2D Mater 8:045026 Schultz T et al (2019) Surface termination dependent work function and electronic properties of Ti 3 C 2 T x MXene. Chem Mater 31:6590–6597 Isari AA, Hayati F, Kakavandi B, Rostami M, Motevassel M, Dehghanifard E, N (2020) Cu co-doped TiO 2 @ functionalized SWCNT photocatalyst coupled with ultrasound and visible-light: an effective sono-photocatalysis process for pharmaceutical wastewaters treatment. Chem Eng J 392:123685 Wu D, Fang H, Lu G, Jiang R, Liu J (2024) Insight into the adsorption and co-adsorption of PPCPs on Ti 3 C 2 T x MXene: behaviors, mechanisms and application potentials. Surf Interfaces 45:103853 Sun T, Li M, Zhou S, Liang M, Chen Y, Zou H (2020) Multi-scale structure construction of carbon fiber surface by electrophoretic deposition and electropolymerization to enhance the interfacial strength of epoxy resin composites. Appl Surf Sci 499:143929 Ghani AA et al (2021) Adsorption and electrochemical regeneration of intercalated Ti3C2Tx MXene for the removal of ciprofloxacin from wastewater. Chem Eng J 421:127780 Thakuria R, Nath NK, Saha BK (2019) The nature and applications of π–π interactions: a perspective. Cryst Growth Des 19:523–528 Li S, Zhang G, Meng D, Yang F (2024) Photoelectrocatalytic activation of sulfate for sulfamethoxazole degradation and simultaneous H 2 production by bifunctional N, P co-doped black-blue TiO2 nanotube array electrode. Chem Eng J 485:149828 Peng Y et al (2021) Charge-transfer resonance and electromagnetic enhancement synergistically enabling MXenes with excellent SERS sensitivity for SARS-CoV-2 S protein detection. Nano Micro Lett 13:52 Caffrey NM (2018) Effect of mixed surface terminations on the structural and electrochemical properties of two-dimensional Ti 3 C 2 T 2 and V 2 CT 2 MXenes multilayers. Nanoscale 10:13520–13530 Klimeš J, Michaelides A (2012) Perspective: Advances and challenges in treating van der Waals dispersion forces in density functional theory. J Chem Phys 137 Additional Declarations There is NO Competing Interest. Supplementary Files SupplementaryMaterial.docx Supplementary Information nrreportingsummary.pdf Cite Share Download PDF Status: Under Review Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7396857","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":590927878,"identity":"c29a8ed3-2ebb-4716-b84c-fddf8e328c5b","order_by":0,"name":"Hong-Guang Guo","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAtUlEQVRIiWNgGAWjYBACAzBZAeFIkKDlDMlaGNtI0WLOv/yaxMd5dfIGB5gP3uZhsMsjqMVyxpsyyZnbDhtuOMCWbM3DkFxM2GE3zqRJ8247kGBwgMdMmofhQGIDcVrm1AG18H8jUsv59mPSvA3MIFvYiLWFh9lyxrHDhjMPsxlbzjFIJsaW4w9vfKipk+c73vzwxpsKO8JaGCRyILHJwAw2gaB6IOA//oAYZaNgFIyCUTCSAQAOrTucFLQdTgAAAABJRU5ErkJggg==","orcid":"","institution":"Sichuan University College of Architecture and Environment","correspondingAuthor":true,"prefix":"","firstName":"Hong-Guang","middleName":"","lastName":"Guo","suffix":""},{"id":590927879,"identity":"daedfa86-fab9-4576-a983-92a5916d5559","order_by":1,"name":"Peng Chen","email":"","orcid":"","institution":"Sichuan University College of Architecture and Environment","correspondingAuthor":false,"prefix":"","firstName":"Peng","middleName":"","lastName":"Chen","suffix":""},{"id":590927880,"identity":"d12549c6-6cc2-43b5-b34d-614a839a6395","order_by":2,"name":"Feiyun Huang","email":"","orcid":"","institution":"Sichuan University college of life science","correspondingAuthor":false,"prefix":"","firstName":"Feiyun","middleName":"","lastName":"Huang","suffix":""},{"id":590927881,"identity":"f13b7dd8-5d8c-4eb6-b697-eabc9c4ea6a4","order_by":3,"name":"Liping Luo","email":"","orcid":"","institution":"Sichuan University College of Architecture and Environment","correspondingAuthor":false,"prefix":"","firstName":"Liping","middleName":"","lastName":"Luo","suffix":""},{"id":590927882,"identity":"05f87b85-3c7d-494b-b66b-cf7553e861fc","order_by":4,"name":"Jingquan Wang","email":"","orcid":"","institution":"Sichuan University","correspondingAuthor":false,"prefix":"","firstName":"Jingquan","middleName":"","lastName":"Wang","suffix":""},{"id":590927883,"identity":"8af35eef-38fe-4f05-9ec6-2b904d29a13d","order_by":5,"name":"Zheng Min","email":"","orcid":"https://orcid.org/0000-0001-9148-7544","institution":"University of New South Wales","correspondingAuthor":false,"prefix":"","firstName":"Zheng","middleName":"","lastName":"Min","suffix":""}],"badges":[],"createdAt":"2025-08-18 07:30:33","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7396857/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7396857/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":102739720,"identity":"15884e8f-144e-4181-80e8-ef3eeb28abd9","added_by":"auto","created_at":"2026-02-16 07:11:34","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":2160189,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eConstruction, microstructure characterization, and surface functionalization of rGAMO-ECs.\u003c/strong\u003e (a-h) Fabrication workflow: (a) Ti₃AlC₂ MAX synthesis, (b) selective HCl/LiF etching to obtain Ti₃C₂Tₓ, (c) delamination into 2D MXene sheets, (d) rGAM composite assembly, (e) in-situ electrooxidation to form rGAMO with surface TiO₂, (f) screen printing of rGAMO electrodes, (g) integration with portable electrochemical workstation, (h) representative I-T response curves. (i-p) TEM and HRTEM characterization: TEM images of (i) rGO, (j) AgNWs, (k) Ti₃C₂Tₓ, and (l) rGO/AgNWs@Ti₃C₂Tₓ; corresponding HRTEM images showing lattice fringes of (m) rGO, (n) AgNWs, (o) Ti₃C₂Tₓ, and (p) rGO/AgNWs@Ti₃C₂Tₓ; insets show SAED patterns. (q) EDS elemental mapping of N, C, Ag, O, and Ti. (r-u) Micro-IR imaging of carbonyl group distribution (1735 cm⁻¹) in (r) Ti₃C₂Tₓ, (s) rGO@Ti₃C₂Tₓ, (t) AgNWs@Ti₃C₂Tₓ, and (u) rGO/AgNWs@Ti₃C₂Tₓ.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7396857/v1/7dba247f7c12a0ad7a408e10.png"},{"id":102739740,"identity":"8d4111c9-0b6e-428d-a8d9-9a1375e29ded","added_by":"auto","created_at":"2026-02-16 07:11:38","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":5825133,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eElectrochemical sensing performance of rGAMO-ECs for CIP and SMX (100-1000 nM). \u003c/strong\u003e(a-b) I-T responses. (c) Calibration curves. (d-e) LSV responses. (f) Polynomial fits. (g) S/N analysis. (h) Baseline noise. (i-k) Optimization of preconcentration time, electrolyte, and scan rate.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7396857/v1/54350ba92427c0135a1d5d77.png"},{"id":102739713,"identity":"f4531ce7-11a6-479c-889e-f27c845f7145","added_by":"auto","created_at":"2026-02-16 07:11:32","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":7316319,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMachine learning-assisted validation of electrochemical fingerprints for antibiotic classification.\u003c/strong\u003e (a) SHAP beeswarm plot of concentration features. (b) Feature importance ranking. (c) t-SNE visualization of CIP, SMX, and mixture samples. (d) ROC curves of XGBoost, Random Forest, and SVM classifiers. (e) Confusion matrix of XGBoost classifier. (f) Learning curves. (g) SHAP waterfall plot of electrochemical features. (h) Feature interaction heatmap.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7396857/v1/7edb03a288e7f809cdbdd9f8.png"},{"id":102739723,"identity":"3d683a5e-7ba4-479c-8297-399edf53dfd9","added_by":"auto","created_at":"2026-02-16 07:11:34","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":3291642,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSelectivity, stability, and reproducibility of rGAMO-ECs.\u003c/strong\u003e (a) Anti-interference tests against target antibiotics, non-target antibiotics, organic molecules, cations, and anions; inset radar plots show response ratios. (b) Long-term stability over 30 days. (c) Batch-to-batch reproducibility across 10 independent batches. Error bars: SD of three parallel measurements.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7396857/v1/48497a3579da1b6ea3c881c2.png"},{"id":102739717,"identity":"d116f090-df61-4d91-83b1-07d2de489cd3","added_by":"auto","created_at":"2026-02-16 07:11:33","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":17210931,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eField validation of rGAMO-ECs in diverse water matrices.\u003c/strong\u003e (a) Sampling region map. (b) Distribution of sampling sites. (c) Dose-response curves and spiked recoveries for CIP and SMX in PBS and real samples (0–1000 nM). (d–f) Screen-printed C-rGAMO, Au-rGAMO, and Pt-rGAMO electrodes. (g–i) On-site testing at Mingyuan Lake (g), Jiang’an River (h), and pond (i).\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7396857/v1/bd8d81a172efaee765839dab.png"},{"id":102739734,"identity":"44787bee-88fc-488e-831a-31a38e0adc87","added_by":"auto","created_at":"2026-02-16 07:11:37","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":847514,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eUnmanned surface vessel equipped with rGAMO-ECs for autonomous long-term monitoring.\u003c/strong\u003e (a) Vessel photograph. (b) Integrated sensing module with real-time data acquisition. (c) Bluetooth-enabled remote monitoring unit. (d) Mobile control terminal. (e) Multi-parameter water quality probe suite.\u003c/p\u003e","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-7396857/v1/b19eb51a4ea4f64378cedc1f.png"},{"id":102739727,"identity":"a359fbcd-3fe5-4ff5-92c1-5677eaab0d8f","added_by":"auto","created_at":"2026-02-16 07:11:36","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":7785076,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDFT-calculated adsorption mechanisms of CIP and SMX on rGAMO-ECs.\u003c/strong\u003e (a-c) Initial, intermediate, and final adsorption configurations. (d) Molecular orbital interaction modes. (e) DOS analysis for pristine and adsorbed systems. (f, h) Potential energy surface landscapes for CIP and SMX. (g, i) Charge density difference maps for CIP-surface and SMX-surface interfaces.\u003c/p\u003e","description":"","filename":"floatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-7396857/v1/3233b20d2a23e0baf36fc396.png"},{"id":104401378,"identity":"62a8ff70-7a43-46f6-a15d-6465348d0080","added_by":"auto","created_at":"2026-03-11 12:12:33","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":46140071,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7396857/v1/e7fa57ca-87ef-4c5e-8f7d-10cd6739ed89.pdf"},{"id":102739737,"identity":"7c937b66-ee82-4e5f-aa05-289e2a719470","added_by":"auto","created_at":"2026-02-16 07:11:38","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":35825702,"visible":true,"origin":"","legend":"Supplementary Information","description":"","filename":"SupplementaryMaterial.docx","url":"https://assets-eu.researchsquare.com/files/rs-7396857/v1/09f225eb71ca23eec1bd4a57.docx"},{"id":102739746,"identity":"9b6cded2-e604-419b-8cea-e70260e2feb9","added_by":"auto","created_at":"2026-02-16 07:11:43","extension":"pdf","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":1671216,"visible":true,"origin":"","legend":"","description":"","filename":"nrreportingsummary.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7396857/v1/47e79964c595b7e1da686e2d.pdf"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Autonomous nanocomposite electrochemical sensing of antibiotics across aquatic ecosystems","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eThe widespread occurrence of antibiotics in aquatic environments is emerging as a major threat to ecosystem integrity and public health, driving the evolution and dissemination of antimicrobial resistance across natural and engineered water systems \u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. Antibiotic contamination occurs through pharmaceutical manufacturing discharge, agricultural runoff, and incomplete wastewater treatment, resulting in persistent environmental concentrations ranging from ng/L to \u0026micro;g/L that foster multidrug-resistant bacteria and threaten ecosystem stability \u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. Current regulatory frameworks mandate maximum allowable concentrations of 0.1\u0026ndash;10 \u0026micro;g/L for different antibiotic classes in surface waters, yet comprehensive monitoring remains severely limited by the lack of portable, field-deployable analytical technologies \u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. Conventional methods such as liquid chromatography-tandem mass spectrometry (LC-MS/MS) provide exceptional sensitivity but require sophisticated instrumentation, specialized laboratory facilities, and extended analysis times, making them impractical for real-time field monitoring \u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003ePortable electrochemical sensors offer compelling advantages, including rapid response within 5 minutes, cost-effectiveness, and miniaturization potential, making them ideal for decentralized environmental surveillance \u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. Research has focused on various nanomaterials for electrode modification, including carbon nanotubes, metal-organic frameworks, noble metal nanoparticles, and two-dimensional materials such as graphene and MXene (where M\u0026thinsp;=\u0026thinsp;Ti, X\u0026thinsp;=\u0026thinsp;C, Tₓ = surface terminations such as -OH, -O, and -F) \u003csup\u003e10, 11, 12\u003c/sup\u003e. Among these, MXene has attracted considerable attention due to its excellent metallic conductivity, abundant surface functional groups, and large specific surface area. To enhance electrochemical performance, researchers have explored hybridization of MXene with metal oxides, metal nanoparticles, and carbon-based materials \u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. However, MXene-based sensors face a critical challenge of oxidation instability in aqueous environments, leading to conductivity degradation that severely limits long-term operational stability and field deployment potential \u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eSeveral strategies have been proposed to address MXene oxidation instability, including surface functionalization, protective coating, and transformation to MXene-metal oxide composites \u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e, \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e. Traditional preparation methods for MXene-TiO₂ composites, such as natural oxidation, high-temperature calcination, and hydrothermal reactions, require extended preparation cycles or extreme conditions, limiting their practical implementation \u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e. Furthermore, field deployment requires sensors capable of operating across diverse water matrices with varying complexity, from simple tap water to environmental samples containing organic matter, suspended particles, and interfering ions, necessitating robust anti-interference capabilities and substrate optimization strategies \u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eHere we establish a new electrochemical materials platform in which the intrinsic oxidation instability of MXenes (Ti₃C₂Tₓ) is converted into a functional phase-engineering strategy, enabling the rapid formation of a conductive Ti₃C₂Tₓ-TiO₂ heterostructure under mild electrochemical conditions. This architecture yields a chemically stable yet electronically active surface that supports multi-analyte electrochemical fingerprinting in complex aqueous environments. To address the challenges of diverse water matrix applications, we further establish a multi-component hierarchical assembly strategy integrating multidimensional nanomaterials, which provides robust anti-interference capabilities and retains potential for bio-adaptive modification to accommodate different environmental conditions. The sensor performance was systematically characterized, and initial differential pulse voltammetry (DPV) screening across five major antibiotic classes (\u003cem\u003eFluoroquinolones\u003c/em\u003e, \u003cem\u003eMacrolides\u003c/em\u003e, \u003cem\u003eTetracyclines\u003c/em\u003e, \u003cem\u003eSulfonamides\u003c/em\u003e, and \u003cem\u003eβ-Lactams\u003c/em\u003e) identified ciprofloxacin (CIP) and sulfamethoxazole (SMX) as exhibiting the most distinctive electrochemical signatures, achieving detection limits of 45 nmol/L and 4.0 nmol/L, respectively. By integrating operando phase-engineered MXenes, electrochemical fingerprinting, machine-learning-based pattern recognition, and molecular-level density functional theory (DFT) analysis, we demonstrate a generalizable electrochemical recognition platform capable of operating across diverse, chemically complex water matrices, from laboratory buffers to real aquatic environments and autonomous field deployment.\u003c/p\u003e"},{"header":"2. Materials and Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Materials Synthesis and Sensor Fabrication\u003c/h2\u003e \u003cdiv id=\"Sec4\" class=\"Section3\"\u003e \u003ch2\u003e2.1.1 Preparation of Ti₃C₂Tₓ nanosheets\u003c/h2\u003e \u003cp\u003eTi₃C₂Tₓ was synthesized via a two-step approach: Ti₃AlC₂ MAX phase (where M\u0026thinsp;=\u0026thinsp;Ti, A\u0026thinsp;=\u0026thinsp;Al, X\u0026thinsp;=\u0026thinsp;C) precursor preparation through ball-milling and high-temperature treatment, followed by selective etching with HCl/LiF and ultrasonic delamination to obtain Ti₃C₂Tₓ nanosheets \u003cb\u003e(details in Supporting Information Text S1 and Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e)\u003c/b\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section3\"\u003e \u003ch2\u003e2.1.2 Fabrication of rGO/AgNWs@Ti₃C₂Tₓ (rGAM) Electrode\u003c/h2\u003e \u003cp\u003eThe rGAM electrode was constructed through hierarchical assembly methodology involving three main components. Graphene oxide (GO) was synthesized via modified Hummer\u0026rsquo;s method, while silver nanowires (AgNWs) were prepared through polyol reduction. The composite assembly was achieved by: (1) solution-phase mixing of GO with AgNWs suspension, (2) screen-printing onto electrode substrate, (3) incorporation of delaminated Ti₃C₂Tₓ nanosheets through controlled dispersion mixing, and (4) electrochemical reduction of GO to reduced graphene oxide (rGO) via cyclic voltammetry (CV) in phosphate-buffered saline (PBS) to form the final rGAM electrode \u003cb\u003e(details in Supporting Information Text S2 and Figure \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e)\u003c/b\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section3\"\u003e \u003ch2\u003e2.1.3 In-situ Electrochemical Oxidation and Electrochemical Sensor (rGAMO-ECs) Fabrication\u003c/h2\u003e \u003cp\u003eThe prepared rGAM electrode underwent in-situ electrochemical oxidation treatment in a custom-designed 50 mL electrochemical reactor equipped with two platinum electrodes (8 mm \u0026times; 8 mm) separated by a 20 mm distance \u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e. The Ti₃C₂Tₓ component was subjected to controlled electrochemical oxidation in 50 mmol/L KCl electrolyte solution under magnetic stirring with pH adjustment and precise electrolyte resistivity control. A constant DC voltage of 6.0 V was systematically applied for various durations (20, 40, 60, 90, and 120 min) while maintaining the reaction temperature at 25\u0026ndash;30\u0026deg;C. During this process, the Ti₃C₂Tₓ component in the rGAM composite was partially oxidized in situ to generate uniformly distributed TiO₂ nanoparticles on the MXene surface, yielding the rGO/AgNWs@Ti₃C₂Tₓ-TiO₂ composite (designated rGAMO). The rGAMO ink was then deposited onto screen-printed three-electrode configurations via screen printing to fabricate the rGAMO-based electrochemical sensor (rGAMO-ECs), which was subsequently integrated with a portable electrochemical workstation for smartphone-assisted data acquisition.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Structural and Morphological Characterization\u003c/h2\u003e \u003cp\u003eComprehensive structural and morphological characterization of rGAMO-ECs composites were performed using multiple complementary techniques. Transmission electron microscopy (TEM), high-resolution TEM (HRTEM), and selected area electron diffraction (SAED) examined morphology, lattice structures, and crystallinity of individual components and composites. Energy-dispersive X-ray spectroscopy (EDS) elemental mapping analyzed spatial distribution of N, C, Ag, O, and Ti elements. Micro-infrared imaging (Micro-IR) mapped carbonyl group distribution (1735 cm⁻\u0026sup1;) across composite interfaces. Scanning electron microscopy (SEM) with elemental mapping characterized surface morphology and elemental distribution. X-ray photoelectron spectroscopy (XPS) determined surface composition and chemical valence states. X-ray diffraction (XRD) assessed crystallographic structure. Fourier-transform infrared spectroscopy (FTIR) identified functional groups. Confocal laser scanning microscopy (CLSM) was used to observe microstructure and interfacial characteristics. Detailed instrumentation parameters and analytical conditions are provided \u003cb\u003ein Supporting Information Text S3\u003c/b\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e2.3 Electrochemical Measurement\u003c/h2\u003e \u003cp\u003eAll electrochemical experiments were performed using a CHI760 electrochemical workstation with a three-electrode system at 25\u0026thinsp;\u0026plusmn;\u0026thinsp;2\u0026deg;C.\u003c/p\u003e \u003cp\u003eFor electrode material optimization (optimal loading), operational stability, anti-interference, and reproducibility tests, rGAM was used as the working electrode, a platinum sheet served as the counter electrode, and a saturated calomel electrode (SCE) was used as the reference electrode. DPV measurements were conducted in 0.1 mol/L PBS (pH 7.6) within a potential range of 0.5\u0026ndash;1.5 V, employing a pulse amplitude of 50 mV and a pulse width of 0.06 s \u003csup\u003e24\u003c/sup\u003e. Electrochemical impedance spectroscopy (EIS) was carried out in a solution containing 5 mM [Fe(CN)\u003csub\u003e6\u003c/sub\u003e]\u003csup\u003e3\u0026minus;/4\u0026minus;\u003c/sup\u003e and 0.1 mol/L KCl over a frequency range from 10⁶ to 0.1 Hz. CV was performed from \u0026minus;\u0026thinsp;0.6 to +\u0026thinsp;0.6 V at scan rates ranging from 20 to 240 mV/s \u003csup\u003e25\u003c/sup\u003e \u003cb\u003e(details in Supporting Information Text S4, Figure S3 and Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e-Table S3)\u003c/b\u003e.\u003c/p\u003e \u003cp\u003eFor precise antibiotic detection and real water sample analysis, rGAMO composite was used as the working electrode, platinum sheet as counter electrode, and SCE as reference electrode. Chronoamperometry (I-T) was performed at +\u0026thinsp;0.95 V vs. SCE for CIP and +\u0026thinsp;0.83 V vs. SCE for SMX, with current responses recorded for 200 s. Linear sweep voltammetry (LSV) was conducted from 0 to +\u0026thinsp;1.4 V vs. SCE at 50 mV/s \u003csup\u003e26\u003c/sup\u003e. All measurements were performed in 50 mM KCl \u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e2.4 Machine Learning (ML) Classification and Model Validation\u003c/h2\u003e \u003cp\u003eThe SHapley Additive exPlanations (SHAP) framework combined with XGBoost was employed for electrochemical feature analysis and antibiotic classification. Six electrochemical parameters were extracted from chronoamperometry (I-T) and linear sweep voltammetry (LSV) measurements at concentrations of 100\u0026ndash;1000 nM. Multi-classifier benchmarking (XGBoost, Random Forest, support vector machine (SVM)) was performed with model evaluation using accuracy, precision, recall, F1-score, and area under the receiver operating characteristic curve (AUC) metrics. t-distributed stochastic neighbor embedding (t-SNE) dimensionality reduction, SHAP analysis, and third-order polynomial regression were applied to elucidate feature contributions and non-linear dependencies \u003cb\u003e(details in Supporting Information Text S5)\u003c/b\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e2.5 Antibiotic Detection and Data Analysis\u003c/h2\u003e \u003cp\u003eStock solutions were prepared in two concentration ranges: 1-1000 nmol/L for precision quantification in environmental samples, and 100\u0026ndash;1000 \u0026micro;mol/L for broad-spectrum screening applications. Peak currents were analyzed using the Randles-Ševč\u0026iacute;k equation with verification of diffusion-controlled processes confirmed through linear Ip vs. ν\u0026sup1;/\u0026sup2; relationships (R\u0026sup2; \u0026gt; 0.99). Detection limits were calculated as limit of detection (LOD) = 3σ/m and limit of quantification (LOQ) = 10σ/m, where σ represents the standard deviation of blank measurements and m is the calibration curve slope \u003cb\u003e(details in Supporting Information Text S6)\u003c/b\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e2.6 Real Sample Testing and Vessel-based Offshore Monitoring\u003c/h2\u003e \u003cp\u003eReal water validation was conducted across multiple aquatic environments including tap water, natural lake water (Mingyuan Lake), flowing river water (Jiang'an River), stagnant pond water, and synthetic urine samples to assess sensor performance under diverse matrix conditions. Water samples were collected using standardized sampling protocols and filtered through 0.45 \u0026micro;m membranes before testing. Synthetic urine was prepared according to established protocols to simulate biological matrix complexity \u003cb\u003e(details in Supporting Information Text S7)\u003c/b\u003e. For large-scale monitoring applications, a single-beam unmanned surface vessel equipped with rGAMO-ECs was designed and deployed for autonomous offshore surveillance.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e2.7 DFT Calculations\u003c/h2\u003e \u003cp\u003eDFT calculations used the Vienna Ab initio Simulation Package (VASP) with the Perdew-Burke-Ernzerhof (PBE) functional under the generalized gradient approximation (GGA) framework, a 500 eV plane-wave cutoff, and a 3\u0026times;3\u0026times;1 Monkhorst-Pack k-point grid \u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e. Van der Waals interactions were described using DFT-D3 dispersion correction \u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e. The composite surface model comprised Ti₃C₂Tₓ MXene with TiO₂ nanoparticles interfaced with rGO layers containing AgNW segments \u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. Geometric optimization was performed until forces\u0026thinsp;\u0026lt;\u0026thinsp;0.02 eV/\u0026Aring; \u003csup\u003e30\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003e2.8 Statistical Analysis\u003c/h2\u003e \u003cp\u003eStatistical analysis employed multiple software platforms including Origin 2021, R software (version 4.2.2), CasaXPS (version 2.3.16), and Python-based machine learning libraries, with significance threshold set at \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05. Detailed methodology for all statistical procedures, machine learning model development, feature importance analysis, and method comparison is provided \u003cb\u003ein Supporting Information Text S8\u003c/b\u003e.\u003c/p\u003e \u003c/div\u003e"},{"header":"3. Results and Discussion","content":"\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003e3.1 Construction and Characterization of rGAMO-ECs\u003c/h2\u003e \u003cp\u003eThe rGAMO-ECs platform was constructed through a hierarchical assembly strategy integrating multidimensional nanomaterials for portable antibiotic sensing at trace levels (~\u0026thinsp;10⁻⁹ mol/L). While Ti₃C₂Tₓ, rGO, and AgNWs have been extensively explored as electrode materials, their synergistic integration through controlled electrooxidation represents a novel approach to address MXene oxidation instability. The fabrication process \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea-h\u003cb\u003e)\u003c/b\u003e begins with Ti₃AlC₂ MAX phase synthesis \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea\u003cb\u003e)\u003c/b\u003e, followed by selective HCl/LiF etching to generate Ti₃C₂Tₓ MXene \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb\u003cb\u003e)\u003c/b\u003e \u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e, \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e. Delamination produces atomically thin nanosheets \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec\u003cb\u003e)\u003c/b\u003e, which are integrated with rGO and AgNWs to form the rGAM composite \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ed\u003cb\u003e)\u003c/b\u003e. The critical transformation occurs through in-situ electrooxidation, generating uniformly distributed TiO₂ nanoparticles while preserving the conductive framework \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ee\u003cb\u003e)\u003c/b\u003e. This approach avoids structural damage from traditional high-temperature treatments, preserving metallic conductivity for rapid electron transfer \u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e and creating a stable core-shell structure that prevents further oxidative degradation \u003csup\u003e\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e. Device fabrication employs screen printing to deposit rGAMO onto three-electrode configurations \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ef\u003cb\u003e)\u003c/b\u003e, with smartphone interfaces for real-time data acquisition \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eg\u003cb\u003e)\u003c/b\u003e. Representative I-T curves show electrochemical responses across trace concentration ranges \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eh\u003cb\u003e)\u003c/b\u003e. This hierarchical assembly simultaneously addresses the restacking tendency of 2D Ti₃C₂Tₓ nanosheets and their well-documented oxidation degradation in aqueous media by intercalating 1D AgNWs as permanent spacer channels and converting the outer Ti₃C₂Tₓ layers into a self-passivating TiO₂ shell via controlled electrochemical oxidation, achieving structural stabilization without the significant conductivity loss (\u0026gt;\u0026thinsp;40%) typically associated with ex situ coating or thermal annealing approaches \u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe hierarchical architecture was characterized via HRTEM \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ei-p\u003cb\u003e)\u003c/b\u003e. The rGO component exhibited wrinkled sheet morphology with interlayer spacing of 0.365\u0026thinsp;\u0026plusmn;\u0026thinsp;0.012 nm \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ei, m\u003cb\u003e)\u003c/b\u003e, corresponding to approximately 68% reduction degree with partial retention of oxygen-containing functional groups serving as molecular anchoring sites \u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e. AgNWs demonstrated a highly crystalline face-centered cubic structure with a uniform diameter of 85\u0026thinsp;\u0026plusmn;\u0026thinsp;15 nm \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ej, n\u003cb\u003e)\u003c/b\u003e, predominantly exposing (111) facets (d-spacing: 0.236\u0026thinsp;\u0026plusmn;\u0026thinsp;0.009 nm, JCPDS #04-0783) that provide optimal π-π stacking interactions with quinolone rings \u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e. Ti₃C₂Tₓ MXene nanosheets displayed an accordion-like layered structure with an interlayer spacing of 0.98\u0026thinsp;\u0026plusmn;\u0026thinsp;0.05 nm \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ek, o\u003cb\u003e)\u003c/b\u003e, with intact Ti-C lattice fringes (d-spacing: 0.245\u0026thinsp;\u0026plusmn;\u0026thinsp;0.012 nm) and sharp hexagonal SAED patterns confirming high crystallinity. The composite structure indicated successful hierarchical assembly with intimate interfacial contact \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003el, p\u003cb\u003e)\u003c/b\u003e. EDS mapping confirmed homogeneous distribution of N, C, Ag, O, and Ti throughout the matrix \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eq\u003cb\u003e)\u003c/b\u003e, with nitrogen originating from polyvinylpyrrolidone (PVP) capping agents and surface amino groups providing hydrogen bonding sites \u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eMicro-IR imaging of carbonyl group distribution (1735 cm⁻\u0026sup1;) revealed surface functionalization patterns \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003er-u\u003cb\u003e)\u003c/b\u003e. Pure Ti₃C₂Tₓ exhibited minimal carbonyl absorption consistent with -OH/-F terminations \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003er\u003cb\u003e)\u003c/b\u003e \u003csup\u003e\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e. Introduction of rGO resulted in heterogeneous carbonyl distribution at sheet edges \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003es\u003cb\u003e)\u003c/b\u003e. The AgNWs@Ti₃C₂Tₓ composite showed enhanced carbonyl density at interfaces \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003et\u003cb\u003e)\u003c/b\u003e. The complete composite displayed the highest and most uniform carbonyl distribution \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eu\u003cb\u003e)\u003c/b\u003e, indicating synergistic creation of oxygen-containing functional groups for antibiotic binding. The uniform carbonyl distribution observed in the ternary rGAMO composite, in contrast to the edge-localized patterns in binary mixtures, indicates that the hierarchical assembly generates emergent interfacial chemical environments serving as hydrogen-bond acceptors for fluoroquinolone moieties and electron-density modulators at TiO₂ active sites for sulfonamide recognition \u003csup\u003e\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eSystematic characterization of electrochemical oxidation determined optimal parameters for TiO₂ nanoparticle generation. CLSM visualization \u003cb\u003e(Figure S4)\u003c/b\u003e revealed progressive TiO₂ formation: sparse fluorescent points at 0\u0026ndash;20 min formed clusters at 20\u0026ndash;60 min, then expanded into interconnected networks at 60\u0026ndash;90 min. The color change from metallic black to gray-black reflects Ti₃C₂Tₓ to TiO₂ transformation \u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e, with preferential oxidation at high-energy sites \u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e, \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e. Fluorescence stabilization after 90 min indicated oxidation equilibrium. SEM analysis \u003cb\u003e(Figure S5)\u003c/b\u003e showed transformation from pristine structures through 10\u0026ndash;20 nm edge nanoparticles to optimal 50\u0026ndash;80 nm networks at 90 min, with over-oxidation at 120 min generating 200\u0026ndash;500 nm aggregates that destroyed conductivity \u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e. Elemental analysis confirmed core-shell structure formation \u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e, \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e, \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e. XPS analysis \u003cb\u003e(Figure S6)\u003c/b\u003e tracked valence evolution: C-Ti bond fluctuation reflects framework reconstruction \u003csup\u003e\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e\u003c/sup\u003e, C-O increase from 12.04% to 15.07% confirms controlled oxidation, and Ti(IV) content peaked at 56.8% at 90 min representing optimal balance. Based on comprehensive characterization, 90-min electrochemical oxidation achieves ideal balance between conductivity and catalytic activity. The well-defined oxidation optimum at 90 min reflects a self-limiting, diffusion-controlled TiO₂ growth mechanism that inherently prevents the runaway oxidation commonly encountered in chemical or thermal routes, providing a practical guideline for reproducible fabrication across different MXene compositions.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003e3.2 Sensing Performance and ML-Assisted Validation of rGAMO-ECs\u003c/h2\u003e \u003cp\u003eInitial DPV screening across five major antibiotic classes (\u003cem\u003eFluoroquinolones, Macrolides, Tetracyclines, Sulfonamides, and β-Lactams\u003c/em\u003e) identified six responsive antibiotics, among which CIP and SMX exhibited the most distinctive electrochemical signatures, prompting their selection for comprehensive investigation \u003cb\u003e(Figure S7)\u003c/b\u003e. Electrochemical I-T and LSV comprehensively evaluated the trace monitoring performance of rGAMO-ECs \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e \u003cb\u003eand Table S4)\u003c/b\u003e. I-T response curves revealed CIP exhibited concentration-dependent characteristics with rapid negative current changes reaching equilibrium within 3\u0026ndash;30 s \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea\u003cb\u003e)\u003c/b\u003e. As CIP concentration increased from 100 to 1000 nM, equilibrium negative response current decreased from \u0026minus;\u0026thinsp;0.045 to -0.005 \u0026micro;A, reflecting cumulative adsorption effects on electrode surfaces \u003csup\u003e\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e, \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e\u003c/sup\u003e. SMX I-T responses \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb\u003cb\u003e)\u003c/b\u003e displayed similar negative trends but with distinctly different kinetics: longer equilibration times (5\u0026ndash;95 s) and larger current changes (-0.17 to -0.02 \u0026micro;A, 0.15 \u0026micro;A total amplitude vs. CIP 0.04 \u0026micro;A). Extended SMX equilibration indicated complex surface interactions involving gradual binding between sulfonamide groups and TiO₂ oxygen vacancies \u003csup\u003e\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e, \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e\u003c/sup\u003e. Quantitative analysis showed CIP exhibited linear response in 100\u0026ndash;1000 nmol/L range with 45 nmol/L detection limit (IEC\u0026thinsp;=\u0026thinsp;1.028\u0026times;10⁻⁴ \u0026times; [CIP]\u0026thinsp;\u0026minus;\u0026thinsp;0.09563, R\u0026sup2; = 0.9344), while SMX demonstrated 4.0 nmol/L detection limit (IEC\u0026thinsp;=\u0026thinsp;2.273\u0026times;10⁻⁵ \u0026times; [SMX]\u0026thinsp;+\u0026thinsp;0.01687, R\u0026sup2; = 0.9120) \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec\u003cb\u003e)\u003c/b\u003e. The 4 nM LOD achieved for SMX is among the lowest reported for non-enzymatic electrochemical detection, approaching chromatographic sensitivity while retaining the advantages of portability and rapid response. The markedly different LODs for CIP (45 nM) and SMX reflect their distinct binding thermodynamics, with SMX benefiting from stronger electrostatic interactions between its sulfonyl group and TiO₂ oxygen vacancies \u003csup\u003e\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eLSV analysis focused on response voltage characteristics at plateau current (100 \u0026micro;A). CIP exhibited voltage decrease with increasing concentration \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ed\u003cb\u003e)\u003c/b\u003e, with response voltages in 1.31\u0026ndash;1.35 V range progressively decreasing from 100 to 1000 nM, indicating catalytic-type electrochemical characteristics that facilitated electron transfer \u003csup\u003e\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e, \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e\u003c/sup\u003e. Conversely, SMX showed voltage increase with concentration \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ee\u003cb\u003e)\u003c/b\u003e, with 1.22\u0026ndash;1.25 V response range displaying upward trends indicating increased electrochemical impedance \u003csup\u003e\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e, \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e\u003c/sup\u003e. Second-order polynomial regression revealed differential responses \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ef\u003cb\u003e)\u003c/b\u003e \u003csup\u003e\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e\u003c/sup\u003e: CIP followed P\u003csub\u003eEV\u003c/sub\u003e (V)\u0026thinsp;=\u0026thinsp;1.3556\u0026ndash;3.045\u0026times;10⁻⁵ \u0026times; [CIP]\u0026thinsp;\u0026minus;\u0026thinsp;3.671\u0026times;10⁻⁸ \u0026times; [CIP]\u0026sup2; (R\u0026sup2; = 0.9924), with negative first-order coefficient indicating surface adsorption facilitating electron transfer \u003csup\u003e\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e\u003c/sup\u003e; SMX followed P\u003csub\u003eEV\u003c/sub\u003e (V)\u0026thinsp;=\u0026thinsp;1.2513\u0026thinsp;+\u0026thinsp;1.157\u0026times;10⁻⁴ \u0026times; [SMX]\u0026thinsp;\u0026minus;\u0026thinsp;6.713\u0026times;10⁻⁸ \u0026times; [SMX]\u0026sup2; (R\u0026sup2; = 0.9811), with positive coefficient reflecting impedance-increasing behavior \u003csup\u003e\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e\u003c/sup\u003e. These distinct mechanisms originated from molecular structure differences: CIP possessed extended π-conjugated systems enabling rapid equilibrium and electron transfer facilitation, while SMX sulfonamide groups exhibited strong TiO₂ affinity but complex surface interactions requiring extended equilibration \u003csup\u003e\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e, \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e62\u003c/span\u003e, \u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e\u003c/sup\u003e. The opposing voltage-concentration trends for CIP and SMX enable a multi-parameter electrochemical fingerprinting approach conceptually analogous to electronic nose arrays \u003csup\u003e\u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e\u003c/sup\u003e, achieved here on a single electrode through the heterogeneous surface chemistry of the ternary nanocomposite, where the extended π-conjugation of CIP facilitates catalytic-type behaviour and the lone-pair donation of SMX to TiO₂ surface states increases interfacial impedance.\u003c/p\u003e \u003cp\u003eSignal quality analysis demonstrated exceptional analytical performance. At 100 nM concentration, signal-to-noise (S/N) ratios reached 52:1 for CIP and 38:1 for SMX \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eg\u003cb\u003e)\u003c/b\u003e, far exceeding detection (S/N\u0026thinsp;=\u0026thinsp;3) and quantification (S/N\u0026thinsp;=\u0026thinsp;10) thresholds \u003csup\u003e\u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e65\u003c/span\u003e\u003c/sup\u003e. Superior signal clarity resulted from rGAMO-enhanced electrocatalytic activity amplifying faradaic currents while continuous conductive networks minimized resistive losses. Baseline noise characterization revealed remarkably low current fluctuations (σ\u0026thinsp;=\u0026thinsp;0.5 nA root-mean-square (RMS), Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eh), representing 4.2-fold reduction versus bare glassy carbon electrodes, attributed to rGO-AgNWs hybrid networks providing parallel conductive pathways and TiO₂ nanoparticles stabilizing interfacial double-layer capacitance \u003csup\u003e\u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e66\u003c/span\u003e\u003c/sup\u003e. Systematic parameter optimization established optimal conditions: preconcentration time studies showed that peak currents reached plateau values at 60 s \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ei\u003cb\u003e)\u003c/b\u003e, achieving 2.8-fold enhancement through spontaneous adsorption driven by π-π stacking and coordination bonding \u003csup\u003e\u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e67\u003c/span\u003e\u003c/sup\u003e; electrolyte optimization identified 50 mM KCl in PBS (pH 7.4) as optimal \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ej\u003cb\u003e)\u003c/b\u003e, providing superior peak currents through phosphate buffering and K⁺ ionic mobility \u003csup\u003e\u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e68\u003c/span\u003e\u003c/sup\u003e; scan rate optimization determined 10 mV/s as ideal \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ek\u003cb\u003e)\u003c/b\u003e, balancing peak resolution (\u0026gt;\u0026thinsp;0.95) and current intensity while minimizing non-faradaic charging current \u003csup\u003e\u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e69\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe rGAMO-ECs successfully established a multi-electrochemical fingerprint recognition system utilizing differences in current change amplitudes, voltage response trends, and reaction kinetic parameters, achieving detection limits of 45 nmol/L for CIP and 4.0 nmol/L for SMX within 100\u0026ndash;1000 nmol/L range. To further validate these electrochemical recognition mechanisms and quantify individual parameter contributions to antibiotic discrimination, comprehensive machine learning analysis combining SHAP interpretation, dimensionality reduction, and multi-classifier validation was employed \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e \u003cb\u003eand Figure S8)\u003c/b\u003e.\u003c/p\u003e \u003cp\u003eSHAP analysis revealed distinct hierarchical patterns in antibiotic discrimination capacity across concentration ranges \u003csup\u003e\u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e70\u003c/span\u003e\u003c/sup\u003e. Low concentrations (100\u0026ndash;400 nM) dominated feature importance, with 100 nM achieving the highest mean absolute SHAP value (0.112), followed by 200 nM (0.104) and 300 nM (0.094) \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea-b\u003cb\u003e)\u003c/b\u003e. This concentration-dependent importance decay reflects sensor saturation kinetics, consistent with Langmuir-Freundlich adsorption models. The bidirectional SHAP distribution demonstrates contrasting electrochemical mechanisms: negative values (-0.3 to -0.1) correspond to CIP's catalytic electron transfer through π-conjugated systems, while positive values (0.1 to 0.3) indicate SMX's impedance-dominated behavior from sulfonamide surface adsorption \u003csup\u003e\u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e71\u003c/span\u003e, \u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e72\u003c/span\u003e\u003c/sup\u003e. Three-dimensional feature space analysis via t-SNE revealed excellent cluster separation between CIP, SMX, and mixture samples \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec\u003cb\u003e)\u003c/b\u003e, validating the discriminative power of electrochemical parameters \u003csup\u003e\u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e73\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eMulti-classifier benchmarking demonstrated exceptional recognition accuracy: XGBoost achieved AUC\u0026thinsp;=\u0026thinsp;0.987, outperforming Random Forest (0.968) and SVM (0.941) \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ed \u003cb\u003eand Table S5)\u003c/b\u003e \u003csup\u003e\u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e74\u003c/span\u003e\u003c/sup\u003e. The XGBoost confusion matrix showed near-perfect classification with only a 2% misclassification rate, where CIP achieved 98% accuracy (147/150), SMX 96.7% (145/150), and mixtures 98.7% (148/150) \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ee \u003cb\u003eand Table S6)\u003c/b\u003e. Learning curve analysis confirmed model generalization without overfitting, with training and validation scores converging at approximately 95% accuracy beyond 300 training samples \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ef\u003cb\u003e)\u003c/b\u003e \u003csup\u003e\u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e75\u003c/span\u003e\u003c/sup\u003e. SHAP waterfall analysis quantified individual electrochemical parameter contributions \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eg\u003cb\u003e)\u003c/b\u003e: I-T Peak current exhibited the highest positive impact (+\u0026thinsp;0.42), followed by LSV equilibration potential (Ep, +\u0026thinsp;0.38), while I-T Area showed negative contribution (-0.25), aligning with theoretical predictions where amperometric peak responses dominate antibiotic discrimination \u003csup\u003e\u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e76\u003c/span\u003e\u003c/sup\u003e. Feature interaction analysis revealed complex non-linear relationships \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eh\u003cb\u003e)\u003c/b\u003e, with strongest interactions between I-T Width and LSV Slope (0.33), suggesting coupled electrochemical processes. In contrast to most ML-assisted sensor studies that treat models as black boxes, our SHAP-based interpretability framework reveals that low-concentration features (100\u0026ndash;400 nM) dominate discrimination capacity within the environmentally relevant range and identifies coupled adsorption-electron transfer processes that remain invisible to single-parameter analysis \u003csup\u003e\u003cspan citationid=\"CR77\" class=\"CitationRef\"\u003e77\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003ePolynomial dependency fitting of electrochemical features \u003cb\u003e(Figure S8 and Table S7)\u003c/b\u003e validated the proposed recognition mechanisms. Third-order polynomial models achieved root-mean-square error (RMSE) values of 1.177\u0026ndash;2.266, with I-T response voltage showing optimal fitting (RMSE\u0026thinsp;=\u0026thinsp;1.177, mean absolute error (MAE)\u0026thinsp;=\u0026thinsp;0.934). The non-monotonic SHAP distributions across voltage ranges confirmed distinct electrochemical fingerprints: SMX exhibited positive potential shifts (1.22\u0026ndash;1.25 V) due to sulfonamide oxidation barriers, while CIP showed negative shifts (1.31\u0026ndash;1.35 V) from fluoroquinolone reduction facilitation \u003csup\u003e\u003cspan citationid=\"CR78\" class=\"CitationRef\"\u003e78\u003c/span\u003e, \u003cspan citationid=\"CR79\" class=\"CitationRef\"\u003e79\u003c/span\u003e\u003c/sup\u003e. The concentration-importance hierarchy (100\u0026ndash;300 nM optimal range) provides crucial design principles for sensor optimization, suggesting measurement precision should prioritize low-concentration detection aligned with environmental monitoring requirements (typical antibiotic levels: 50\u0026ndash;500 nM). These machine learning results collectively validate the electrochemical recognition mechanisms of rGAMO-ECs and confirm its reliable performance for trace-level antibiotic monitoring.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003e3.3 Selectivity, Stability, and Reproducibility of rGAMO-ECs\u003c/h2\u003e \u003cp\u003eAnti-interference performance evaluation \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea\u003cb\u003e)\u003c/b\u003e demonstrated that the rGAMO-ECs exhibited exceptional selective recognition capability for target antibiotics CIP and SMX. The sensor achieved response voltages of 1.33 V for CIP and 1.25 V for SMX, significantly higher than responses to interferents. Importantly, the sensor showed extremely low cross-responses to common water interferents including interfering antibiotics (ampicillin, amoxicillin, tetracycline, erythromycin, chloramphenicol), small organic molecules (glucose, urea, ascorbic acid, dopamine, serotonin), and inorganic ions (Ca\u0026sup2;⁺, Mg\u0026sup2;⁺, Fe\u0026sup3;⁺, Al\u0026sup3;⁺, Cl⁻, SO₄\u0026sup2;⁻, NO₃⁻, PO₄\u0026sup3;⁻, CO₃\u0026sup2;⁻, HCO₃⁻), with response voltages maintained in the low 1.00-1.08 V range. This excellent anti-interference performance stemmed from the synergistic multi-recognition site mechanism constructed on the composite material surface \u003csup\u003e\u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e80\u003c/span\u003e, \u003cspan citationid=\"CR81\" class=\"CitationRef\"\u003e81\u003c/span\u003e\u003c/sup\u003e. The ability to discriminate CIP and SMX from their structural analogues (e.g., enrofloxacin and sulfadiazine) arises from the multi-site recognition mechanism of the ternary nanocomposite, in which rGO π-domains, AgNW facets, and TiO₂ oxygen vacancy sites collectively encode molecular geometry rather than responding to a single functional group, thereby achieving lock-and-key-like selectivity without the limitations inherent to molecularly imprinted polymers \u003csup\u003e\u003cspan citationid=\"CR82\" class=\"CitationRef\"\u003e82\u003c/span\u003e\u003c/sup\u003e. Long-term stability testing \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eb\u003cb\u003e)\u003c/b\u003e validated the outstanding durability of the rGAMO-ECs sensor. During 30 days of continuous monitoring, the sensor maintained response stability of 98.5% for CIP and 98.4% for SMX, significantly superior to rGAM-ECs (95.3% and 95.7%) and rGM-ECs (92.6% and 94.2%) sensors. The exceptional long-term stability primarily resulted from TiO₂ nanoparticle introduction, which formed stable Ti-O-C and Ti-O-Ag chemical bonding networks at composite interfaces, effectively suppressing AgNW oxidative degradation during long-term electrochemical cycling while protecting MXene substrates from structural damage \u003csup\u003e\u003cspan citationid=\"CR83\" class=\"CitationRef\"\u003e83\u003c/span\u003e, \u003cspan citationid=\"CR84\" class=\"CitationRef\"\u003e84\u003c/span\u003e\u003c/sup\u003e. The \u0026gt;\u0026thinsp;98% signal retention over 30 days substantially exceeds the 7\u0026ndash;14 day operational lifetimes typical of MXene-based sensors \u003csup\u003e\u003cspan citationid=\"CR85\" class=\"CitationRef\"\u003e85\u003c/span\u003e\u003c/sup\u003e, and comparative data across electrode configurations confirm that the in situ electrochemical oxidation step, rather than the mere multi-component assembly, is the primary contributor to long-term durability through formation of a self-passivating TiO₂ barrier layer. Batch-to-batch reproducibility was evaluated across 10 independently fabricated sensor batches \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ec\u003cb\u003e)\u003c/b\u003e. Response voltages were 1.324\u0026ndash;1.347 V for CIP with relative standard deviation (RSD) of 0.63%, and 1.239\u0026ndash;1.268 V for SMX with RSD of 0.67%. The assessment followed the IUPAC single-laboratory validation framework and the Eurachem fitness-for-purpose principle, with the Association of Official Analytical Chemists (AOAC)/Horwitz model used to contextualize acceptable precision at trace levels. Within this framework, the inter-batch RSDs indicated robust process stability and reproducibility, supporting subsequent scale-up and practical implementation \u003csup\u003e\u003cspan citationid=\"CR86\" class=\"CitationRef\"\u003e86\u003c/span\u003e, \u003cspan citationid=\"CR87\" class=\"CitationRef\"\u003e87\u003c/span\u003e\u003c/sup\u003e. These results indicated that the established sensor preparation method possessed good process stability and reproducibility, providing a solid foundation for subsequent scale-up and practical implementation. The \u0026lt;\u0026thinsp;1% inter-batch RSDs are competitive with commercial screen-printed electrodes (2\u0026ndash;5% RSD) and significantly outperform most laboratory-fabricated nanocomposite sensors (5\u0026ndash;15% RSD) \u003csup\u003e\u003cspan citationid=\"CR88\" class=\"CitationRef\"\u003e88\u003c/span\u003e\u003c/sup\u003e, which can be attributed to the self-regulating nature of electrochemical oxidation that minimizes batch-to-batch variability in the critical surface activation step.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003e3.4 Field Validation of rGAMO-ECs in Diverse Water Matrices\u003c/h2\u003e \u003cp\u003eTo validate the practical applicability of rGAMO-ECs in real-world environments, various actual samples representing different application scenarios were systematically evaluated using the spiked recovery method \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e \u003cb\u003eand Figures S9)\u003c/b\u003e. Field testing was conducted across multiple sampling sites \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ea-b\u003cb\u003e)\u003c/b\u003e, including tap water, Mingyuan Lake, Jiang'an River, pond water, and artificial urine, covering matrix types ranging from simple to complex with varying flow conditions. Portable sensing apparatus deployment (\u003cb\u003eFigure S9a, d, g\u003c/b\u003e) demonstrated the system's field adaptability across diverse aquatic environments. The complete setup included high-power AC/DC mobile power supply, membraneless air pump, filtration apparatus, portable electrochemical workstation with mobile phone interface, and reaction chamber, enabling real-time on-site detection. Considering the different flow characteristics and complexity of water environments, three electrode substrates were prepared: C-rGAMO electrodes \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ed \u003cb\u003eand Figure S9c)\u003c/b\u003e for moderately flowing calm lake water, Au-rGAMO electrodes \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ee \u003cb\u003eand Figure S9f)\u003c/b\u003e for flowing systems such as rivers and tap water, and Pt-rGAMO electrodes \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ef \u003cb\u003eand Figure S9i)\u003c/b\u003e for nearly stagnant environments like pond water rich in organic matter (algae, etc.) and synthetic urine samples. This substrate selection approach optimizes sensor performance according to flow dynamics and matrix characteristics, where carbon substrates are suitable for moderate flow conditions, gold substrates provide enhanced performance in dynamic flowing systems, and platinum substrates exhibit superior anti-fouling capabilities in stagnant, organically complex environments \u003csup\u003e\u003cspan citationid=\"CR89\" class=\"CitationRef\"\u003e89\u003c/span\u003e\u003c/sup\u003e. The sensor demonstrated good analytical performance across all tested matrices \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ec\u003cb\u003e)\u003c/b\u003e, with recovery rates maintained within the acceptable range of 85\u0026ndash;115%. Recovery rates for CIP detection ranged from 86.4% to 113.7%, while SMX detection achieved recovery rates of 85.1% to 109.8%. Dose-response curves maintained good linearity (R\u0026sup2; \u0026gt; 0.99) across the entire concentration range (0-1000 nM). Comparative analysis of filtration membranes \u003cb\u003e(Figure S9j)\u003c/b\u003e revealed distinct color variations reflecting differences in organic matter (algae) and sediment content across water sources. The mobile phone control interface \u003cb\u003e(Figure S9k)\u003c/b\u003e enabled real-time monitoring and data acquisition during I-T electrochemical testing, demonstrating the system's user-friendly operation and practical deployment capabilities. This substrate optimization approach aligns with current trends in portable electrochemical sensor development and helps address matrix interference, a common challenge faced by biosensors \u003csup\u003e\u003cspan citationid=\"CR90\" class=\"CitationRef\"\u003e90\u003c/span\u003e, \u003cspan citationid=\"CR91\" class=\"CitationRef\"\u003e91\u003c/span\u003e\u003c/sup\u003e. The matrix-adaptive substrate selection strategy addresses a common limitation in sensor validation where performance is assessed in only a single matrix type, and the 85\u0026ndash;115% recovery range achieved across all matrices meets the acceptance criteria established by US EPA Method for pharmaceutical residue analysis in environmental waters \u003csup\u003e\u003cspan citationid=\"CR92\" class=\"CitationRef\"\u003e92\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003ePerformance analysis of different water sources indicated that sensor response was closely related to matrix complexity and substrate selection. Tap water, as the simplest matrix with relatively stable ionic composition and fewer interferents, yielded the most consistent results using C-rGAMO electrodes (CIP: 94.2%-108.6%; SMX: 91.8%-106.3%). Environmental water samples from Mingyuan Lake and Jiang'an River tested with Au-rGAMO electrodes showed moderate matrix effects while maintaining acceptable analytical precision. Pond water, containing higher concentrations of organic matter, suspended particles, and various ions, represented the most complex environmental matrix and required Pt-rGAMO electrodes to achieve acceptable recovery rates (CIP: 86.4%-113.7%; SMX: 85.1%-108.9%). Artificial urine samples showed intermediate analytical performance (CIP: 93.1%-109.4%; SMX: 90.7%-107.5%), indicating good biocompatibility and anti-protein fouling capabilities of the sensor. Field application testing \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eg-i\u003cb\u003e)\u003c/b\u003e validated the practical feasibility of this approach, with the portable workstation combined with optimized electrodes enabling on-site detection capabilities for various water bodies. By matching electrode complexity to matrix requirements, this detection method maintains analytical performance while controlling costs, meeting the needs for online monitoring of pollutants in complex environments.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003e3.5 Vessel-Mounted rGAMO-ECs for Long-Term Autonomous Monitoring\u003c/h2\u003e \u003cp\u003eTo validate the application potential of rGAMO-ECs in large-scale water monitoring, a single-beam unmanned surface vessel equipped with rGAMO-ECs was designed and constructed \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e\u003cb\u003e)\u003c/b\u003e. The vessel-based monitoring system \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ea-e\u003cb\u003e)\u003c/b\u003e integrated solar power, Global Positioning System (GPS) positioning, fourth-generation (4G) data transmission, and flow-through detection chamber, achieving 72-hour continuous autonomous operation covering\u0026thinsp;\u0026gt;\u0026thinsp;10 km\u0026sup2; area. The integration of chemical-specific trace sensors with autonomous surface vessels addresses a critical gap in environmental monitoring where fixed stations offer limited spatial coverage and grab sampling fails to capture contaminant plume dynamics \u003csup\u003e\u003cspan citationid=\"CR93\" class=\"CitationRef\"\u003e93\u003c/span\u003e\u003c/sup\u003e. The successful 72-hour continuous operation confirms that the rGAMO-ECs platform meets the stability and power efficiency requirements for unattended field deployment.\u003c/p\u003e \u003cp\u003eField validation through parallel LC-MS/MS analysis confirmed the vessel-based system's analytical reliability \u003cb\u003e(Figure S10 and Table S8)\u003c/b\u003e. LC-MS/MS total ion chromatogram \u003cb\u003e(Figure S10a)\u003c/b\u003e showed complete separation of SMX (3.42 min) and CIP (4.78 min) from internal standards SMX-d4 (3.38 min) and CIP-d8 (4.73 min), with peak intensities reaching 8.24\u0026times;10⁵ and 1.18\u0026times;10⁶ cps, respectively. MS/MS spectra confirmed molecular structures: CIP \u003cb\u003e(Figure S10b)\u003c/b\u003e produced characteristic fragment ions m/z 314.1 (base peak, 94.8%), 288.1 (62.3%), and 245.0 (44.7%) at collision energy 15 eV; SMX \u003cb\u003e(Figure S10c)\u003c/b\u003e yielded m/z 156.0 (86.7%), 108.0 (71.2%), and 92.0 (54.8%) at 18 eV. Calibration curves \u003cb\u003e(Figure S10d)\u003c/b\u003e demonstrated excellent linearity across 0.3\u0026ndash;5000 nmol/L: CIP followed y=1847x\u0026thinsp;+\u0026thinsp;232 (R\u0026sup2;=0.998), SMX followed y=2157x\u0026thinsp;+\u0026thinsp;184 (R\u0026sup2;=0.997). Method comparison \u003cb\u003e(Figure S10e)\u003c/b\u003e showed strong correlation between rGAMO-ECs and LC-MS/MS (y\u0026thinsp;=\u0026thinsp;0.962x\u0026thinsp;+\u0026thinsp;3.84, R\u0026sup2;=0.987). Bland-Altman analysis \u003cb\u003e(Figure S10f)\u003c/b\u003e confirmed minimal systematic bias with mean difference\u0026thinsp;\u0026minus;\u0026thinsp;11.6 nmol/L and 95% limits of agreement\u0026thinsp;\u0026plusmn;\u0026thinsp;48.7 nmol/L.\u003c/p\u003e \u003cp\u003eThe strong LC-MS/MS correlation (R\u0026sup2; = 0.987) and 95% limits of agreement (\u0026plusmn;\u0026thinsp;48.7 nmol/L) within the regulatory-relevant concentration range provide rigorous statistical evidence that the vessel-mounted rGAMO-ECs can serve as a reliable field surrogate for laboratory chromatographic methods \u003csup\u003e\u003cspan citationid=\"CR94\" class=\"CitationRef\"\u003e94\u003c/span\u003e\u003c/sup\u003e, supporting reduced dependence on centralized analytical infrastructure.\u003c/p\u003e \u003cp\u003eReal aquaculture effluent validation \u003cb\u003e(Table S9)\u003c/b\u003e demonstrated sensor performance under extreme conditions. Fish pond effluents (Site A) achieved CIP recovery of 83.6\u0026ndash;86.4% and SMX recovery of 81.4\u0026ndash;86.3%, despite high concentrations of interfering antibiotics (oxytetracycline 45.3 nmol/L, tetracycline 28.7 nmol/L, sulfamethazine 62.4 nmol/L, enrofloxacin 15.8 nmol/L). Shrimp pond effluents (Site B) showed similar recoveries with different interferent profiles (sulfadiazine 78.2 nmol/L, sulfathiazole 34.5 nmol/L, norfloxacin 41.3 nmol/L). Matrix effects ranged from \u0026minus;\u0026thinsp;10.7% to -13.7%, primarily due to high organic loads (chemical oxygen demand (COD): 186\u0026ndash;342 mg/L, five-day biochemical oxygen demand (BOD₅): 95\u0026ndash;168 mg/L, dissolved organic carbon (DOC): 52.8\u0026ndash;97.3 mg/L). Mixed sample simultaneous detection validated multiplexing capability: 500 nmol/L CIP and SMX mixed spikes achieved individual recoveries of 82.4% and 80.6%, respectively, with total recovery of 81.5%. Detection limits in matrices increased compared to pure water (CIP from 45 to 68.5 nmol/L, SMX from 4.0 to 8.32 nmol/L), but remained well below regulatory limits. Despite the exceptionally challenging matrix conditions of aquaculture effluents, the maintained recoveries above 80% and successful simultaneous CIP/SMX detection validate the orthogonality of the electrochemical fingerprints for resolving individual analytes in mixed contamination scenarios typical of aquaculture discharge \u003csup\u003e\u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e95\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe vessel-based system maintained stable performance under complex water quality conditions: pH 7.2\u0026ndash;8.4, temperature 25\u0026ndash;28\u0026deg;C, dissolved oxygen 3.2\u0026ndash;5.8 mg/L, ammonia nitrogen 12.5\u0026ndash;35.7 mg/L. These parameters represent typical intensive aquaculture conditions, validating sensor reliability in real application scenarios. Cost-benefit analysis revealed that the vessel system with monthly sensor replacement is significantly more economical than traditional sampling and LC-MS/MS analysis, while providing real-time data (\u0026lt;\u0026thinsp;5 minutes) versus laboratory analysis (24\u0026ndash;48 hours). The tolerance to wide-ranging water quality parameters without recalibration reflects the buffering capacity of the multi-component electrode architecture, and when scaled to regional vessel networks, this approach could provide the continuous, spatially resolved antibiotic surveillance data required for evidence-based antimicrobial resistance mitigation policies \u003csup\u003e\u003cspan citationid=\"CR96\" class=\"CitationRef\"\u003e96\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003e3.6 Theoretical Elucidation of Antibiotic-Specific Recognition Mechanisms on rGAMO-ECs\u003c/h2\u003e \u003cp\u003eMulti-stage adsorption configuration evolution of CIP and SMX molecules on rGAMO-ECs composite surfaces revealed complete mechanistic pathways from initial contact to final stable adsorption \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e, \u003cb\u003eTable S10-S11)\u003c/b\u003e. In the initial stage \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ea\u003cb\u003e)\u003c/b\u003e, CIP fluorine atoms preferentially established electrostatic interactions with cationic vacancies on Ti₃C₂Tₓ surfaces, achieving initial binding energy of -1.24 eV (F: -0.78 eV, O: -0.46 eV) \u003csup\u003e\u003cspan citationid=\"CR97\" class=\"CitationRef\"\u003e97\u003c/span\u003e, \u003cspan citationid=\"CR98\" class=\"CitationRef\"\u003e98\u003c/span\u003e\u003c/sup\u003e. SMX initial recognition occurred through sulfonyl oxygen interactions with coordinatively unsaturated titanium sites on TiO₂, yielding stronger initial binding energy of -1.67 eV (sulfonyl O: -1.21 eV, oxazole ring: -0.46 eV) \u003csup\u003e\u003cspan citationid=\"CR99\" class=\"CitationRef\"\u003e99\u003c/span\u003e\u003c/sup\u003e. During adsorption progression \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eb\u003cb\u003e)\u003c/b\u003e, molecules underwent significant spatial rearrangement: CIP rotated\u0026thinsp;~\u0026thinsp;23\u0026deg; enabling oxygen atoms to establish additional interactions forming dual-point anchoring (binding energy\u0026thinsp;\u0026minus;\u0026thinsp;1.89 eV: F -0.89 eV, O -1.00 eV) \u003csup\u003e\u003cspan citationid=\"CR100\" class=\"CitationRef\"\u003e100\u003c/span\u003e\u003c/sup\u003e, while SMX experienced\u0026thinsp;~\u0026thinsp;31\u0026deg; orientation adjustment with sulfur atoms becoming active centers (binding energy\u0026thinsp;\u0026minus;\u0026thinsp;2.15 eV: sulfonyl O -1.35 eV, S -0.80 eV). In fully optimized final configurations \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ec\u003cb\u003e)\u003c/b\u003e, CIP achieved \"full-contact\" mode through further\u0026thinsp;~\u0026thinsp;15\u0026deg; rotation (total\u0026thinsp;~\u0026thinsp;38\u0026deg;), forming multi-point anchoring networks with maximum binding energy of -2.47 eV (F: -1.02 eV, O: -1.45 eV) \u003csup\u003e\u003cspan citationid=\"CR100\" class=\"CitationRef\"\u003e100\u003c/span\u003e, \u003cspan citationid=\"CR101\" class=\"CitationRef\"\u003e101\u003c/span\u003e, \u003cspan citationid=\"CR102\" class=\"CitationRef\"\u003e102\u003c/span\u003e\u003c/sup\u003e. SMX final configuration (total rotation\u0026thinsp;~\u0026thinsp;45\u0026deg;) displayed complex three-point anchoring involving sulfonyl sulfur, sulfonyl oxygen, and oxazole ring oxygen, achieving optimal binding energy of -2.83 eV (sulfonyl O: -1.47 eV, S: -0.94 eV, oxazole O: -0.42 eV).\u003c/p\u003e \u003cp\u003eMolecular orbital analysis revealed fundamental differences in electronic interactions \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ed\u003cb\u003e)\u003c/b\u003e. CIP exhibited two binding modes: \"attraction-dominated\" binding involving π orbital overlap with rGO (-0.73 eV) and σ-type interactions with Ti₃C₂Tₓ/TiO₂, and \"competitive equilibrium\" binding with balanced attractive/repulsive orbital interactions (total covalent contributions: -1.12 eV) \u003csup\u003e\u003cspan citationid=\"CR103\" class=\"CitationRef\"\u003e103\u003c/span\u003e\u003c/sup\u003e. SMX exhibited \"dual-side tight attraction\" characteristics with sulfonyl lone-pair orbitals forming donor-acceptor interactions with Ti₃C₂Tₓ d orbitals and oxazole π systems mixing with TiO₂ oxygen p orbitals, resulting in enhanced covalent contributions of -1.47 eV (Ti-O bonds: -0.62 eV, AgNWs coordination: -0.61 eV) \u003csup\u003e\u003cspan citationid=\"CR104\" class=\"CitationRef\"\u003e104\u003c/span\u003e\u003c/sup\u003e. Density of states (DOS) analysis \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ee\u003cb\u003e)\u003c/b\u003e showed pristine CIP and SMX with characteristic peaks at 0.5 eV and 0.8 eV above Fermi level. Upon adsorption, CIP-surface system generated new states at -2.5 eV and \u0026minus;\u0026thinsp;1.2 eV below the Fermi level (EF) (integrated DOS: 3.2 states/eV), indicating strong σ-bonding, while SMX introduced broader state distribution between \u0026minus;\u0026thinsp;3.0 and \u0026minus;\u0026thinsp;1.0 eV (integrated DOS: 4.1 states/eV), confirming enhanced orbital hybridization and stronger electronic coupling. The π-dominated binding of CIP through its conjugated quinolone system creates efficient charge transfer channels that manifest as a voltage-decreasing response, whereas the lone-pair-mediated donor-acceptor binding of SMX with Ti d-orbitals introduces localized electronic states that increase interfacial impedance and produce a voltage-rising response, representing two fundamentally different transduction pathways coexisting on the same heterogeneous electrode surface \u003csup\u003e\u003cspan citationid=\"CR105\" class=\"CitationRef\"\u003e105\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003ePotential energy surface landscapes (PES) provided critical thermodynamic and kinetic insights. For CIP \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ef\u003cb\u003e)\u003c/b\u003e, the PES revealed primary energy well at coordinates (0, -1) \u0026Aring; corresponding to O-Ti interaction sites, with barrier height 1.08 eV and well depth\u0026thinsp;\u0026minus;\u0026thinsp;2.47 eV \u003csup\u003e106\u003c/sup\u003e. Energy gradient along the approach pathway measured\u0026thinsp;\u0026minus;\u0026thinsp;0.36 eV/\u0026Aring; with characteristic funnel radius 2.8 \u0026Aring; facilitating molecular orientation. Secondary local minima at (\u0026plusmn;\u0026thinsp;2, 0) \u0026Aring; (depth\u0026thinsp;\u0026minus;\u0026thinsp;0.72 eV) represented alternative F-Ti binding sites. SMX PES \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eh\u003cb\u003e)\u003c/b\u003e displayed a broader energy basin with primary minimum at (0, 0) \u0026Aring; reaching\u0026thinsp;\u0026minus;\u0026thinsp;2.83 eV depth, and multiple secondary minima at (\u0026plusmn;\u0026thinsp;2, \u0026plusmn;\u0026thinsp;1) \u0026Aring; (-1.54 eV depth) corresponding to Ring-N and SO-O coordination sites. A gentler approach gradient (-0.24 eV/\u0026Aring;) and extended flat region (radius\u0026thinsp;~\u0026thinsp;3.5 \u0026Aring;) around primary minimum indicated lower kinetic barriers (0.85 eV) but greater configurational flexibility.\u003c/p\u003e \u003cp\u003eCharge density difference analysis revealed distinct electronic redistribution patterns at molecule-surface interfaces. For CIP \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eg\u003cb\u003e)\u003c/b\u003e, charge accumulation regions (Δρ = +0.36\u0026nbsp;e/\u0026Aring;\u0026sup3;) concentrated between F atoms and Ti sites at 2.1 \u0026Aring; distances, with complementary depletion zones (Δρ = -0.48\u0026nbsp;e/\u0026Aring;\u0026sup3;) surrounding quinolone backbone. Integrated charge transfer amounted to 0.42 electrons from CIP to surface with 2.3 Debye dipole moment change. SMX charge density mapping \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ei\u003cb\u003e)\u003c/b\u003e showed more pronounced polarization with intense accumulation at sulfonyl-Ti interfaces (Δρ = +0.24\u0026nbsp;e/\u0026Aring;\u0026sup3;) at 1.9 \u0026Aring; and significant depletion around benzene ring (Δρ = -0.52\u0026nbsp;e/\u0026Aring;\u0026sup3;). A net transfer of 0.58 electrons and 3.1 Debye dipole moment change confirmed stronger ionic character in SMX binding. Comprehensive interaction energy analysis demonstrated electrostatic interactions contributed more to SMX binding (-0.89 eV) versus CIP (-0.52 eV), primarily through stronger S-Ti binding (-0.54 eV) versus F-Ti interactions (-0.31 eV). Van der Waals forces favored CIP binding (-0.51 eV versus SMX\u0026thinsp;\u0026minus;\u0026thinsp;0.31 eV), while induced dipole interactions showed greater significance for CIP (-0.32 eV) than SMX (-0.16 eV), reflecting different electronic environments and polarization characteristics \u003csup\u003e\u003cspan citationid=\"CR107\" class=\"CitationRef\"\u003e107\u003c/span\u003e\u003c/sup\u003e. These quantitative findings aligned with experimental adsorption isotherms showing SMX maximum capacity of 287.3 mg/g exceeding CIP 243.7 mg/g, validating stronger theoretical binding energy predictions.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"4. Conclusions","content":"\u003cp\u003eThis work demonstrates that the intrinsic oxidation instability of MXenes can be harnessed as a phase-engineering strategy to create stable, catalytically active sensing interfaces. The rGAMO-ECs platform achieves nanomolar detection limits (45 nmol/L for CIP and 4.0 nmol/L for SMX), \u0026gt;\u0026thinsp;98% signal retention over 30 days, and 96.7% machine-learning classification accuracy, while DFT calculations reveal fundamentally different attraction-dominated (CIP) and impedance-dominated (SMX) recognition mechanisms that provide molecular-level design principles for tailoring electrode surfaces. Field validation confirms strong LC-MS/MS correlation (R\u0026sup2; = 0.987) across diverse water matrices, and autonomous vessel deployment enables 72-hour continuous monitoring covering\u0026thinsp;\u0026gt;\u0026thinsp;10 km\u0026sup2;, further demonstrating that nanocomposite electrochemical sensors can transition from laboratory settings to large-scale environmental surveillance.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgments\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors would like to thank the National Key Research and Development Program of China (2023YFC3210100), National Natural Science Foundation of China (42177060), and Science \u0026amp; Technology Department of Sichuan (2023NSFSC1949, 2023ZHCGO024-LH) for the financial support.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eLoffler P, Escher B, Baduel C, Virta MP, Lai FY (2023) Antimicrobial transformation products in the aquatic environment: global occurrence, ecotoxicological risks, and potential of antibiotic resistance. Environ Sci Technol 57:9474\u0026ndash;9494\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRodriguez-Mozaz S et al (2020) Antibiotic residues in final effluents of European wastewater treatment plants and their impact on the aquatic environment. Environ Int 140:105733\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang Q, Ying G, Pan C, Liu Y-S, Zhao J-L (2015) Comprehensive evaluation of antibiotics emission and fate in the river basins of China: source analysis, multimedia modeling, and linkage to bacterial resistance. Environ Sci Technol 49:6772\u0026ndash;6782\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHanna N et al (2018) Presence of antibiotic residues in various environmental compartments of Shandong province in eastern China: its potential for resistance development and ecological and human risk. Environ Int 114:131\u0026ndash;142\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiguori K et al (2022) Antimicrobial resistance monitoring of water environments: a framework for standardized methods and quality control. Environ Sci Technol 56:9149\u0026ndash;9160\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDorival-Garc\u0026iacute;a, Zafra‐G\u0026oacute;mez C, V\u0026iacute;lchez Naval\u0026oacute;n (2013) Simultaneous determination of 13 quinolone antibiotic derivatives in wastewater samples using solid‐phase extraction and ultra performance liquid chromatography\u0026ndash;tandem mass spectrometry. Microchem J 106:323\u0026ndash;333\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFrigoli M et al (2024) Electrochemical Sensors for Antibiotic Detection: A Focused Review with a Brief Overview of Commercial Technologies. Sensors 24:5576\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi R et al (2020) A flexible and physically transient electrochemical sensor for real-time wireless nitric oxide monitoring. Nat Commun 11:3207\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCoatsworth P et al (2024) Time-resolved chemical monitoring of whole plant roots with printed electrochemical sensors and machine learning. Sci Adv 10:6315\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWu HB, Lou XW (2017) Metal-organic frameworks and their derived materials for electrochemical energy storage and conversion: Promises and challenges. Sci Adv 3:9252\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGao C, Yu X-Y, Xu R-X, Liu J-H, Huang X-J (2012) AlOOH-reduced graphene oxide nanocomposites: one-pot hydrothermal synthesis and their enhanced electrochemical activity for heavy metal ions. ACS Appl Mater Interfaces 4:4672\u0026ndash;4682\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang R, Wang S, Zhang Y, Jin D, Tao X, Zhang L (2018) Graphene-coupled Ti\u003csub\u003e3\u003c/sub\u003eC\u003csub\u003e2\u003c/sub\u003e MXenes-derived TiO\u003csub\u003e2\u003c/sub\u003e mesostructure: promising sodium-ion capacitor anode with fast ion storage and long-term cycling. J Mater Chem A 6:1017\u0026ndash;1027\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHantanasirisakul K, Gogotsi Y (2018) Electronic and optical properties of 2D transition metal carbides and nitrides (MXenes). Adv Mater 30:1804779\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAnasori B, Lukatskaya MR, Gogotsi Y (2023) 2D metal carbides and nitrides (MXenes) for energy storage. \u003cem\u003eMXenes\u003c/em\u003e). Jenny Stanford Publishing\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNaguib M et al (2023) Two-dimensional nanocrystals produced by exfoliation of Ti\u003csub\u003e3\u003c/sub\u003eAlC\u003csub\u003e2\u003c/sub\u003e. \u003cem\u003eMXenes\u003c/em\u003e). Jenny Stanford Publishing\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYang Z et al (2024) Nanozyme-enhanced electrochemical biosensors: mechanisms and applications. Small 20:2307815\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTian W et al (2019) Multifunctional nanocomposites with high strength and capacitance using 2D MXene and 1D nanocellulose. Adv Mater 31:1902977\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNayak P, Kurra N, Xia C, Alshareef HN (2016) Highly efficient laser scribed graphene electrodes for on-chip electrochemical sensing applications. Adv Electron Mater 2:1600185\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePazniak H et al (2020) Partially oxidized Ti3C2T x MXenes for fast and selective detection of organic vapors at part-per-million concentrations. ACS Appl Nano Mater 3:3195\u0026ndash;3204\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGopinath KP, Madhav NV, Krishnan A, Malolan R, Rangarajan G (2020) Present applications of titanium dioxide for the photocatalytic removal of pollutants from water: A review. J Environ Manage 270:110906\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eReddy YVM, Sravani B, Łuczak T, Mallikarjuna K, Madhavi G (2021) An ultra-sensitive rifampicin electrochemical sensor based on titanium nanoparticles (TiO\u003csub\u003e2)\u003c/sub\u003e anchored reduced graphene oxide modified glassy carbon electrode. Colloids Surf Physicochem Eng Aspects 608:125533\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRhouati A, Berkani M, Vasseghian Y, Golzadeh N (2022) MXene-based electrochemical sensors for detection of environmental pollutants: A comprehensive review. Chemosphere 291:132921\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKovalenko I et al (2011) A major constituent of brown algae for use in high-capacity Li-ion batteries. Science 334:75\u0026ndash;79\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSharma N, Selvam SP, Yun K (2020) Electrochemical detection of amikacin sulphate using reduced graphene oxide and silver nanoparticles nanocomposite. Appl Surf Sci 512:145742\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi Y et al (2020) A highly sensitive and selective molecularly imprinted electrochemical sensor modified with TiO\u003csub\u003e2\u003c/sub\u003e-reduced graphene oxide nanocomposite for determination of podophyllotoxin in real samples. J Electroanal Chem 873:114439\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLuo J et al (2017) Pillared structure design of MXene with ultralarge interlayer spacing for high-performance lithium-ion capacitors. ACS Nano 11:2459\u0026ndash;2469\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTurlybekuly A, Shynybekov Y, Soltabayev B, Yergaliuly G, Mentbayeva A (2024) The cross-sensitivity of chemiresistive gas sensors: Nature, methods, and peculiarities: A systematic review. ACS Sens 9:6358\u0026ndash;6371\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKresse G, Furthm\u0026uuml;ller J (1996) Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Phys Rev B 54:11169\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGrimme S, Antony J, Ehrlich S, Krieg H (2010) A consistent and accurate ab initio parametrization of density functional dispersion correction (DFT-D) for the 94 elements H-Pu. J Chem Phys 132\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHenkelman G, Uberuaga BP, J\u0026oacute;nsson H (2000) A climbing image nudged elastic band method for finding saddle points and minimum energy paths. J Chem Phys 113:9901\u0026ndash;9904\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen J et al (2020) Recent progress and advances in the environmental applications of MXene related materials. Nanoscale 12:3574\u0026ndash;3592\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKim Y-J et al (2021) Etching mechanism of monoatomic aluminum layers during MXene synthesis. Chem Mater 33:6346\u0026ndash;6355\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIqbal A, Hong J, Ko TY, Koo CM (2021) Improving oxidation stability of 2D MXenes: synthesis, storage media, and conditions. Nano Convergence 8:9\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSoomro RA, Zhang P, Fan B, Wei Y, Xu B (2023) Progression in the oxidation stability of MXenes. Nano Micro Lett 15:108\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNatu V, Clites M, Pomerantseva E (2018) Mesoporous MXene powders synthesized by acid induced crumpling and their use as Na-ion battery anodes. Mater Res Lett 6:230\u0026ndash;235\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDreyer DR, Park S, Bielawski CW, Ruoff RS (2010) The chemistry of graphene oxide. Chem Soc Rev 39:228\u0026ndash;240\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLazar P et al (2013) Adsorption of small organic molecules on graphene. J Am Chem Soc 135:6372\u0026ndash;6377\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang CJ et al (2017) Oxidation stability of colloidal two-dimensional titanium carbides (MXenes). Chem Mater 29:4848\u0026ndash;4856\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHope MA et al (2016) NMR reveals the surface functionalisation of Ti\u003csub\u003e3\u003c/sub\u003eC\u003csub\u003e2\u003c/sub\u003e MXene. Phys Chem Chem Phys 18:5099\u0026ndash;5102\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSeh ZW, Kibsgaard J, Dickens C (2017) Combining theory and experiment in electrocatalysis: Insights into materials design. Science 355:eaad4998\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTran NM, Ta QTH, Noh J-S (2021) Unusual synthesis of safflower-shaped TiO\u003csub\u003e2\u003c/sub\u003e/Ti\u003csub\u003e3\u003c/sub\u003eC\u003csub\u003e2\u003c/sub\u003e heterostructures initiated from two-dimensional Ti\u003csub\u003e3\u003c/sub\u003eC\u003csub\u003e2\u003c/sub\u003e MXene. Appl Surf Sci 538:148023\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHe S, Shang C, Lu H, Xu Z, Zhao H (2025) Experimental and numerical study of TiO\u003csub\u003e2\u003c/sub\u003e nanoparticle evolution in a diffusion flame reactor. Combust Flame 273:113965\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLu Y et al (2018) Self-hydrogenated shell promoting photocatalytic H\u003csub\u003e2\u003c/sub\u003e evolution on anatase TiO\u003csub\u003e2\u003c/sub\u003e. Nat Commun 9:2752\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShang H et al (2023) Surface hydrogen bond-induced oxygen vacancies of TiO\u003csub\u003e2\u003c/sub\u003e for two-electron molecular oxygen activation and efficient NO oxidation. Environ Sci Technol 57:20400\u0026ndash;20409\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGhassemi H et al (2014) In situ environmental transmission electron microscopy study of oxidation of two-dimensional Ti \u003csub\u003e3\u003c/sub\u003e C \u003csub\u003e2\u003c/sub\u003e and formation of carbon-supported TiO \u003csub\u003e2\u003c/sub\u003e. J Mater Chem A 2:14339\u0026ndash;14343\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNoman M et al (2024) Ti3C2Tx-MXene based 2D/3D Ti\u003csub\u003e3\u003c/sub\u003eC\u003csub\u003e2\u003c/sub\u003e\u0026ndash;TiO\u003csub\u003e2\u003c/sub\u003e\u0026ndash;CuTiO\u003csub\u003e3\u003c/sub\u003e heterostructure for enhanced pseudocapacitive performance. Chem Eng J 499:156697\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXia Y et al (2018) Thickness-independent capacitance of vertically aligned liquid-crystalline MXenes. Nature 557:409\u0026ndash;412\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu N et al (2022) High-temperature stability in air of Ti\u003csub\u003e3\u003c/sub\u003eC\u003csub\u003e2\u003c/sub\u003eT\u003csub\u003ex\u003c/sub\u003e MXene-based composite with extracted bentonite. Nat Commun 13:5551\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhou J, Huang S, He Z, Song S (2021) Enhanced activity and stability of PbO2 electrodes by modification with octadecyl phosphonic acid. J Electrochem Soc 168:116503\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDo BV et al (2023) Voltammetric determination of a fluoroquinolone antibiotic based on multilayer reduced graphene oxide sensor prepared directly, promptly by electrochemically expanding graphite electrode surface. Int J Environ Anal Chem 103:8141\u0026ndash;8157\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLim J, Yang Y, Hoffmann MR (2019) Activation of peroxymonosulfate by oxygen vacancies-enriched cobalt-doped black TiO\u003csub\u003e2\u003c/sub\u003e nanotubes for the removal of organic pollutants. Environ Sci Technol 53:6972\u0026ndash;6980\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCao Y et al (2020) Fabrication of novel CuFe\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e4\u003c/sub\u003e/MXene hierarchical heterostructures for enhanced photocatalytic degradation of sulfonamides under visible light. J Hazard Mater 387:122021\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMajdinasab M, Daneshi M, Marty JL (2021) Recent developments in non-enzymatic (bio) sensors for detection of pesticide residues: Focusing on antibody, aptamer and molecularly imprinted polymer. Talanta 232:122397\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShafiei H, Hassaninejad-Darzi SK (2023) Electroanalytical application of Ag@ POM@ rGO nanocomposite and ionic liquid modified carbon paste electrode for the quantification of ciprofloxacin antibiotic. J Electroanal Chem 935:117321\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSantos AM, Wong A, Almeida AA, Fatibello-Filho O (2017) Simultaneous determination of paracetamol and ciprofloxacin in biological fluid samples using a glassy carbon electrode modified with graphene oxide and nickel oxide nanoparticles. Talanta 174:610\u0026ndash;618\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhu X et al (2022) The effect of sulfamethoxazole on nitrogen removal and electricity generation in a tidal flow constructed wetland coupled with a microbial fuel cell system: Microbial response. Chem Eng J 431:134070\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang S, Song H-L, Yang X-L, Li H, Wang Y-W (2018) A system composed of a biofilm electrode reactor and a microbial fuel cell-constructed wetland exhibited efficient sulfamethoxazole removal but induced sul genes. Bioresour Technol 256:224\u0026ndash;231\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXu X, Wu S, Guo D, Niu X (2020) Construction of a recyclable oxidase-mimicking Fe3O4@ MnOx-based colorimetric sensor array for quantifying and identifying chlorophenols. Anal Chim Acta 1107:203\u0026ndash;212\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi S, Zhang X, Huang Y (2017) Zeolitic imidazolate framework-8 derived nanoporous carbon as an effective and recyclable adsorbent for removal of ciprofloxacin antibiotics from water. J Hazard Mater 321:711\u0026ndash;719\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang S, Zhang J, Gharbi O, Vivier V, Gao M, Orazem ME (2021) Electrochemical impedance spectroscopy. Nat Rev Methods Primers 1:41\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZheng X et al (2024) Oxygen vacancies-promoted removal of COS via catalytic hydrolysis over CuTiO2-δ nanoflowers. Chem Eng J 492:152322\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhu L, Santiago-Sch\u0026uuml;bel B, Xiao H, Hollert H, Kueppers S (2016) Electrochemical oxidation of fluoroquinolone antibiotics: Mechanism, residual antibacterial activity and toxicity change. Water Res 102:52\u0026ndash;62\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZeng W, Zhang H, Wu R, Liu L, Li G, Liang H (2023) Environment-friendly and efficient electrochemical degradation of sulfamethoxazole using reduced TiO\u003csub\u003e2\u003c/sub\u003e nanotube arrays-based Ti membrane coated with Sb-SnO\u003csub\u003e2\u003c/sub\u003e. J Hazard Mater 446:130642\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePeveler W, Yazdani M, Rotello V (2016) Selectivity and specificity: pros and cons in sensing. ACS Sens 1:1282\u0026ndash;1285\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMaduraiveeran G, Sasidharan M, Ganesan V (2018) Electrochemical sensor and biosensor platforms based on advanced nanomaterials for biological and biomedical applications. Biosens Bioelectron 103:113\u0026ndash;129\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang X, Chen Y, Schmidt OG, Yan C (2016) Engineered nanomembranes for smart energy storage devices. Chem Soc Rev 45:1308\u0026ndash;1330\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSun Y, He K, Zhang Z, Zhou A, Duan H (2015) Real-time electrochemical detection of hydrogen peroxide secretion in live cells by Pt nanoparticles decorated graphene\u0026ndash;carbon nanotube hybrid paper electrode. Biosens Bioelectron 68:358\u0026ndash;364\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNi Y, Yue W, Liu F, Bi W, Sun Z, Wu Y (2023) Efficient electrochemical oxidation of cephalosporin antibiotics by a highly active cerium doped PbO\u003csub\u003e2\u003c/sub\u003e anode: Parameters optimization, kinetics and degradation pathways. Colloids Surf Physicochem Eng Aspects 666:131318\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMartin-Yerga D, Costa-Garcia A, Unwin PR (2019) Correlative voltammetric microscopy: structure\u0026ndash;activity relationships in the microscopic electrochemical behavior of screen printed carbon electrodes. ACS Sens 4:2173\u0026ndash;2180\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLundberg SM, Lee S-I (2017) A unified approach to interpreting model predictions. Adv Neural Inf Process Syst 30\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGayen P, Chaplin BP (2016) Selective electrochemical detection of ciprofloxacin with a porous nafion/multiwalled carbon nanotube composite film electrode. ACS Appl Mater Interfaces 8:1615\u0026ndash;1626\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLabib M, Sargent EH, Kelley SO (2016) Electrochemical methods for the analysis of clinically relevant biomolecules. Chem Rev 116:9001\u0026ndash;9090\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMaaten Lvd, Hinton G (2008) Visualizing data using t-SNE. J Mach Learn Res 9:2579\u0026ndash;2605\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEster M, Kriegel H, Xu X, XGBoost: (2016) A scalable tree boosting system. In Proceedings of the 22Nd ACM SIGKDD International Conference on Knowledge Discovery and Data Miningvol, pg 785, \u003cem\u003eGeogr Anal\u003c/em\u003e, (2022)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCui F, Yue Y, Zhang Y, Zhang Z, Zhou HS (2020) Advancing biosensors with machine learning. ACS Sens 5:3346\u0026ndash;3364\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePuthongkham P, Wirojsaengthong S, Suea-Ngam A (2021) Machine learning and chemometrics for electrochemical sensors: moving forward to the future of analytical chemistry. Analyst 146:6351\u0026ndash;6364\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWei L, Chen Z, Shi L (2017) Super-multiplex vibrational imaging. Nature 544:465\u0026ndash;470\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNguyen T-B et al (2023) Phosphoric acid-activated biochar derived from sunflower seed husk: Selective antibiotic adsorption behavior and mechanism. Bioresour Technol 371:128593\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIgwegbe CA, Oba SN, Aniagor CO, Adeniyi AG, Ighalo JO (2021) Adsorption of ciprofloxacin from water: a comprehensive review. J Ind Eng Chem 93:57\u0026ndash;77\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen K, Liu S, Zhou Y, Zhao G (2025) Monitoring and Analysis of Total Tetracyclines in Water from Different Environmental Scenarios Using a Designed Broad-Spectrum Aptamer. Environ Sci Technol 59:5736\u0026ndash;5746\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuang W et al (2025) Wearable Sensor for Continuous Monitoring Multiple Biofluids: Improved Performances by Conductive Metal-Organic Framework with Dual‐Redox Sites on Flexible Graphene Fiber Microelectrode. Adv Funct Mater, 2424018\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBelBruno JJ (2018) Molecularly imprinted polymers. Chem Rev 119:94\u0026ndash;119\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen J et al (2023) Reverse oxygen spillover triggered by CO adsorption on Sn-doped Pt/TiO\u003csub\u003e2\u003c/sub\u003e for low-temperature CO oxidation. Nat Commun 14:3477\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBhat A, Anwer S, Bhat KS, Mohideen MIH, Liao K, Qurashi A (2021) Prospects challenges and stability of 2D MXenes for clean energy conversion and storage applications. npj 2D Mater Appl 5:61\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSoomro RA, Jawaid S, Zhu Q, Abbas Z, Xu B (2020) A mini-review on MXenes as versatile substrate for advanced sensors. Chin Chem Lett 31:922\u0026ndash;930\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBorman PJ et al (2024) Ongoing analytical procedure performance verification using a risk-based approach to determine performance monitoring requirements. Anal Chem 96:966\u0026ndash;979\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHern\u0026aacute;ndez F, Fabregat-Safont D, Campos-Ma\u0026ntilde;as M, Quintana JB (2023) Efficient validation strategies in environmental analytical chemistry: A focus on organic micropollutants in water samples. Annu Rev Anal Chem 16:401\u0026ndash;428\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHayat A, Marty JL (2014) Disposable screen printed electrochemical sensors: Tools for environmental monitoring. Sensors 14:10432\u0026ndash;10453\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBunnasit S, Thamsirianunt K, Rakthabut R, Jeamjumnunja K, Prasittichai C, Siriwatcharapiboon W (2024) Sensitive portable electrochemical sensors for antibiotic chloramphenicol by tin/reduced graphene oxide-modified screen-printed carbon electrodes. ACS Appl Nano Mater 7:267\u0026ndash;278\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSilva FWL et al (2024) Disposable electrochemical sensor: Highly sensitive determination of nitrofurazone antibiotic in environmental samples and pharmaceutical formulations. Chemosphere 361:142481\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKhezerloo E, Hekmat F, Zad AI, Shahrokhian S (2025) A novel electrochemical probe crafted from copper oxide Nanoball-multiwalled carbon nanotube for valacyclovir detection in environmental and medicinal matrices. Electrochim Acta, 146872\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMao K, Zhang H, Pan Y (2021) Biosensors for wastewater-based epidemiology for monitoring public health. Water Res 191:116787\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDunbabin M, Marques L (2012) Robots for environmental monitoring: Significant advancements and applications. IEEE Rob Autom Magazine 19:24\u0026ndash;39\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThompson M, Ellison SL, Wood R (2002) Harmonized guidelines for single-laboratory validation of methods of analysis (IUPAC Technical Report). \u003cem\u003ePure Appl Chem\u003c/em\u003e 74, 835\u0026ndash;855\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLulijwa R, Rupia EJ, Alfaro AC (2020) Antibiotic use in aquaculture, policies and regulation, health and environmental risks: a review of the top 15 major producers. Rev Aquacult 12:640\u0026ndash;663\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLarsson DJ, Flach C-F (2022) Antibiotic resistance in the environment. Nat Rev Microbiol 20:257\u0026ndash;269\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eN\u0026auml;slund L-\u0026Aring;, Mikkel\u0026auml; M-H, Kokkonen E, Magnuson M (2021) Chemical bonding of termination species in 2D carbides investigated through valence band UPS/XPS of Ti\u003csub\u003e3\u003c/sub\u003eC\u003csub\u003e2\u003c/sub\u003eT\u003csub\u003ex\u003c/sub\u003e MXene. 2D Mater 8:045026\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchultz T et al (2019) Surface termination dependent work function and electronic properties of Ti\u003csub\u003e3\u003c/sub\u003eC\u003csub\u003e2\u003c/sub\u003eT\u003csub\u003ex\u003c/sub\u003e MXene. Chem Mater 31:6590\u0026ndash;6597\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIsari AA, Hayati F, Kakavandi B, Rostami M, Motevassel M, Dehghanifard E, N (2020) Cu co-doped TiO\u003csub\u003e2\u003c/sub\u003e@ functionalized SWCNT photocatalyst coupled with ultrasound and visible-light: an effective sono-photocatalysis process for pharmaceutical wastewaters treatment. Chem Eng J 392:123685\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWu D, Fang H, Lu G, Jiang R, Liu J (2024) Insight into the adsorption and co-adsorption of PPCPs on Ti\u003csub\u003e3\u003c/sub\u003eC\u003csub\u003e2\u003c/sub\u003eT\u003csub\u003ex\u003c/sub\u003e MXene: behaviors, mechanisms and application potentials. Surf Interfaces 45:103853\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSun T, Li M, Zhou S, Liang M, Chen Y, Zou H (2020) Multi-scale structure construction of carbon fiber surface by electrophoretic deposition and electropolymerization to enhance the interfacial strength of epoxy resin composites. Appl Surf Sci 499:143929\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGhani AA et al (2021) Adsorption and electrochemical regeneration of intercalated Ti3C2Tx MXene for the removal of ciprofloxacin from wastewater. Chem Eng J 421:127780\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThakuria R, Nath NK, Saha BK (2019) The nature and applications of π\u0026ndash;π interactions: a perspective. Cryst Growth Des 19:523\u0026ndash;528\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi S, Zhang G, Meng D, Yang F (2024) Photoelectrocatalytic activation of sulfate for sulfamethoxazole degradation and simultaneous H\u003csub\u003e2\u003c/sub\u003e production by bifunctional N, P co-doped black-blue TiO2 nanotube array electrode. Chem Eng J 485:149828\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePeng Y et al (2021) Charge-transfer resonance and electromagnetic enhancement synergistically enabling MXenes with excellent SERS sensitivity for SARS-CoV-2 S protein detection. Nano Micro Lett 13:52\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCaffrey NM (2018) Effect of mixed surface terminations on the structural and electrochemical properties of two-dimensional Ti\u003csub\u003e3\u003c/sub\u003eC\u003csub\u003e2\u003c/sub\u003eT\u003csub\u003e2\u003c/sub\u003e and V\u003csub\u003e2\u003c/sub\u003eCT\u003csub\u003e2\u003c/sub\u003e MXenes multilayers. Nanoscale 10:13520\u0026ndash;13530\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKlimeš J, Michaelides A (2012) Perspective: Advances and challenges in treating van der Waals dispersion forces in density functional theory. J Chem Phys 137\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"MXene-TiO₂ nanocomposite, Portable electrochemical sensor, Machine learning classification, Unmanned vessel monitoring, Aquatic antibiotic detection","lastPublishedDoi":"10.21203/rs.3.rs-7396857/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7396857/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eAntibiotics are pervasive in aquatic environments, driving antimicrobial resistance and threatening ecosystem and human health, yet monitoring still depends on laboratory-based methods that preclude real-time, large-scale surveillance. Here we report an autonomous electrochemical sensing platform based on a stabilized MXene (Ti₃C₂Tₓ) nanocomposite that enables selective, ultrasensitive, and long-term monitoring of antibiotics directly in natural waters. A controlled in situ electrochemical oxidation strategy converts Ti₃C₂Tₓ into a Ti₃C₂Tₓ-TiO₂ heterostructure while preserving electrical conductivity, producing a chemically robust and catalytically active sensing interface. When integrated with reduced graphene oxide and silver nanowires, the resulting hierarchical electrode generates multi-site molecular recognition that yields distinct electrochemical fingerprints for ciprofloxacin and sulfamethoxazole. The sensor achieves nanomolar detection limits (45 nM for ciprofloxacin and 4 nM for sulfamethoxazole), maintains\u0026thinsp;\u0026gt;\u0026thinsp;98% signal retention over 30 days, and discriminates target antibiotics from structurally related compounds and complex matrix interferents. Machine-learning analysis of time- and voltage-resolved electrochemical features enables 96.7% classification accuracy, while density functional theory reveals fundamentally different adsorption and charge-transfer mechanisms for the two antibiotics. Field validation across lakes, rivers, aquaculture effluents, and synthetic biological fluids shows strong agreement with LC-MS/MS measurements (R\u003csup\u003e2\u003c/sup\u003e\u0026thinsp;=\u0026thinsp;0.987). When deployed on an unmanned surface vessel, the system enables continuous, autonomous monitoring over \u0026gt;\u0026thinsp;10 km\u003csup\u003e2\u003c/sup\u003e for 72 hours. Together, these results establish electrochemical fingerprinting on stabilized MXene nanocomposites as a scalable strategy for real-time surveillance of antibiotic pollution in aquatic ecosystems.\u003c/p\u003e","manuscriptTitle":"Autonomous nanocomposite electrochemical sensing of antibiotics across aquatic ecosystems","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-02-16 07:10:27","doi":"10.21203/rs.3.rs-7396857/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-communications","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"NCOMMS","sideBox":"Learn more about [Nature Communications](http://www.nature.com/ncomms/)","snPcode":"","submissionUrl":"https://mts-ncomms.nature.com/","title":"Nature Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Communications","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"7c1d4777-2561-4ca5-9ca0-97ab9aae9ec9","owner":[],"postedDate":"February 16th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":62887475,"name":"Earth and environmental sciences/Environmental sciences/Environmental chemistry/Environmental monitoring"},{"id":62887476,"name":"Earth and environmental sciences/Environmental sciences/Environmental impact"}],"tags":[],"updatedAt":"2026-04-21T13:01:37+00:00","versionOfRecord":[],"versionCreatedAt":"2026-02-16 07:10:27","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7396857","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7396857","identity":"rs-7396857","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}