Experimental

Reagents and materials

ERL was received as a gift from a pharmaceutical company (Hyderabad, India). TiO2 nanoparticles (anatase, ≥99 %, particle size < 30 nm) were procured from (CAS No: 13463-67-7, NCZD2902, Nanochemazone, Canada). A 1 mM ERL stock solution was prepared in methanol and stored at 4 °C in a dark container. All reagents were of analytical reagent grade. Phosphate buffer (PB, 0.1 M, pH 7.4) was prepared using KH2PO4 and Na2HPO4 in deionized (DI) water and used as the supporting electrolyte throughout. Screen-printed electrodes were fabricated using conductive carbon ink (Code No: 050-Sun Chemical C2030519P4, Bangalore), conductive silver ink (Code No: 060; Siltech Corporation Inc., Bangalore), and conductive Ag/AgCl paste (CAS No: 7783-90-06, NCZ-SPI-104, Nanochemazone, Canada) on 25 μm polyethylene terephthalate (PET) sheets.

Instrumentation

Electrochemical experiments were performed using a DY2300 Bipotentiostat (Digi Ivy, USA). SEM-EDAX analysis was performed using a Gemini SEM 300 (Carl Zeiss, Germany). XRD was recorded using an Aeris Research benchtop X-ray diffractometer (Malvern PANalytical, UK). SPEs were directly analysed by FTIR spectroscopy using Shimadzu IRSprit-X (Serial no. A230962). EIS was performed using the CHI6005E Electrochemical Workstation of Central Instrumentation Facility of IIT Palakkad.

Fabrication of the TiO₂ nanoparticles-modified screen-printed electrode

Bare SPEs were fabricated by screen-printing a silver conductive layer followed by a carbon layer and an Ag/AgCl reference electrode layer on PET sheets, as described in our previous study [39]. For surface modification, TiO₂NPs were thoroughly mixed with carbon paste at a weight ratio of 1:4 (TiO₂NPs : carbon paste) to obtain a homogeneous composite. The resulting paste was then screen-printed over the working electrode surface and subsequently dried at 60 °C to ensure complete solvent evaporation and firm adhesion of the composite layer. The modified electrode (TiO₂NP @SPE) was then conditioned by running 10 consecutive cyclic voltammograms in 0.1 M PB (pH 7.4) prior to use. Conditioning was performed to stabilize the electrode surface and remove loosely adsorbed nanoparticles, ensuring reproducible baseline responses.

Electrochemical characterization

The electrochemically active area of the TiO2NP@SPE was determined using cyclic voltammetry in 5 mM [Fe(CN)6]3-/4- with 0.1 M KCl at scan rates ranging from 2 to 200 mV s-1, applying the Randles-Ševčík equation. Scan rate studies for ERL oxidation were conducted with 15 μM ERL in 0.1 M PB (pH 7.4) at scan rates from 5 to 200 mV s-1 (potential window: 0.2-0.9 V). The nature of the electrode process (adsorption vs. diffusion-controlled) was determined from plots of _I_pa vs. v and _I_pa vs. v-1/2. The supporting electrolyte pH was maintained at 7.4 throughout, as this value corresponds to the physiological pH of human serum and was established as the optimal condition for ERL oxidation at SPE in our previous study [39]; since TiO₂NP modification does not alter the fundamental redox mechanism of ERL as confirmed by the identical irreversible oxidation behaviour and consistent peak potential observed in the scan rate study and electrode kinetics below so a separate pH optimization on the modified electrode was not performed

Sensitivity studies and calibration

DPV measurements were performed in 0.1 M PB (pH 7.4, as described in the Results and discussion, section Sensitivity studies: calibration and detection limits) over a potential window of 0 to 0.8 V, with ERL concentrations ranging from 15 to 65 μM. DPV parameters were: pulse amplitude 50 mV, potential increment 50 mV, pulse width 50 ms, and pulse period 100 ms, which were optimized in previous experiments [39]; the parameters that afforded the maximum peak current and best peak resolution were selected. The limit of detection (LOD) and limit of quantification (LOQ) were calculated as 3_σ_/m and 10_σ_/m, respectively, where σ is the standard deviation of the current response of 10 replicate measurements of the blank (0.1 M PB, pH 7.4, in the absence of ERL) and m is the slope of the calibration curve.

Selectivity studies

The selectivity of the TiO2NP@SPE was evaluated in the presence of common biological interferents includeing inorganic salts (Mg2⁺, Na2⁺, K⁺, Ca2⁺), sugars (glucose, sucrose, maltose), dopamine, urea, uric acid, fluoroquinolones (moxifloxacin (MOX), levofloxacin (LEV)), antifungal drug tinidazole (TNZ), acetaminophen (ACT), pantoprazole (PNP), metformin (MET) and structurally similar anticancer drugs (imatinib (IMT), capecitabine (CAP), sunitinib (STB), dasatinib (DSB)). Voltammetric responses were recorded in the presence of each interferent at concentrations that mimic physiological conditions, together with 22 μM ERL. Interferent concentrations were selected to reflect physiological levels in human serum or to represent a 1000-fold excess over the ERL concentration tested, whichever was higher.

Reproducibility, stability and recovery studies

Repeatability was assessed by comparing the DPV responses of 8 runs on a single modified electrode with 55 μM ERL. Reproducibility was assessed by comparing the DPV response of 6 individually prepared TiO2NP@SPE electrodes to 55 μM ERL. Stability was evaluated over 30 days on electrodes stored at room temperature. Recovery studies were conducted on two human serum samples (obtained from healthy volunteers) spiked with ERL at three concentrations (33, 44 and 55 μM) using the standard addition method. Both intra-day (n = 6) and inter-day analyses were performed. Student's t-test (95 % confidence, degrees of freedom (df = 5)) was applied to assess statistical significance.

Statistical data analysis

Experimental data were collected in triplicate and analysed using the Microsoft Excel Data Analysis ToolPak. One-way analysis of variance (ANOVA) in Origin 2026b [40] was used to determine whether the ERL concentrations identified using the developed approach differed significantly.