Integrated computational and experimental design of copper–gallic acid nanozymes for selective salbutamol detection

Abstract

Salbutamol (SBM), a β₂-adrenergic agonist commonly prescribed for bronchospasm, is increasingly monitored in sports medicine due to its potential misuse as a performance-enhancing agent. To address the need for rapid, cost-effective, and portable anti-doping diagnostics, a nanozyme-based sensing platform using copper–gallic acid hybrid structures (Cu@GA·HSs) was developed. These nanozymes exhibited oxidase-like activity, enabling sensitive and selective optical detection of SBM across a wide concentration range (125–5000 μg mL⁻¹), with excellent analytical accuracy (96.8–99.8%) and precision (CV < 2.0%). The system demonstrated strong operational stability, retaining 60% catalytic activity after 24 days at 4 °C, and strong resistance to chemical interference, including metal ions and non-polar solvents (selectivity coefficient ≈1). Benchmarking against HPLC revealed excellent agreement (R² = 0.997; deviation <0.06%), validating its analytical performance. Molecular docking and dynamics simulations further revealed specific SBM–matrix interactions underlying the sensor's selectivity and robustness. Together, these results highlight a systems-level approach integrating nanozyme chemistry with computational modeling to engineer next-generation biosensors for anti-doping applications.

---

Design, System, Application

A copper–gallic acid (Cu@GA) hybrid nanozyme was engineered via a green, self-assembly route to mimic oxidase-like activity under mild conditions. The molecular design focused on coordinating Cu²⁺ ions with gallic acid ligands to form a hybrid system capable of catalyzing chromogenic reactions without requiring H₂O₂. This approach provides a low-cost, scalable, and environmentally friendly alternative to conventional enzyme-based systems. The sensor's functionality was optimized by tuning the metal-to-ligand ratio to maximize catalytic stability and selectivity toward β₂-agonists. The sensing system combines stable catalytic performance, robustness against solvent and ion interference, and reliable colorimetric response in a wide concentration range. Molecular docking and dynamics simulations revealed moderately dynamic, site-specific interactions between salbutamol and the hybrid surface, offering insights into binding behavior and signal reproducibility. This work establishes a multifunctional, non-biological catalyst system suitable for selective analyte detection in complex environments. It offers immediate relevance to anti-doping and clinical diagnostics, where on-site, instrument-free, and selective detection of performance-enhancing substances is critical. The modularity of the metal–phenolic coordination framework enables future adaptation for detecting other small-molecule targets, expanding its utility in portable diagnostics and sensor technologies.

1. Introduction

Salbutamol (SBM), a short-acting and selective beta-2 agonist, is commonly used to prevent and treat bronchospasm in disorders such as asthma and chronic obstructive pulmonary disease.¹ It is typically administered via inhalation to directly affect the beta-2-adrenergic receptors as a sympathomimetic, commonly prescribed to counteract bronchoconstriction.² Many athletes believe this mechanism could enhance their performance by relaxing smooth muscles in the lungs and potentially exerting anabolic effects on skeletal muscles.³ Despite its therapeutic uses, some athletes therefore associate SBM with doping.⁴ Consequently, the World Anti-Doping Agency (WADA) has limited the maximum daily dosage of this beta-2 agonist.⁵

Detecting and identifying SBM in athletes not only preserves the integrity of sports but also protects the health of athletes.⁴ Therefore, continuous research in developing robust detection methods is crucial for combating doping issues. Various techniques have been explored for detecting SBM, including high-performance liquid chromatography (HPLC), thin-layer chromatography, spectrophotometric, spectrofluorimetric methods, and immunoassays.⁶ Each of these methods has its strengths and weaknesses. For example, chromatographic methods provide high accuracy but can be time-consuming and require intricate sample preparation.⁶ Voltammetry methods may have some analytical errors and require careful environmental control to minimize toxicity. Moreover, flow-injection spectrophotometry consumes many reagents.⁷ Spectroscopic methods offer several benefits over other methods, including ease of use, cost-effectiveness, rapid detection, and availability of a wide range of optical probes.⁸ Optical sensors, inexpensive tools, have shown excellent performance in detecting doping compounds.⁸ Furthermore, the ability of this class of sensors to provide a visible response allows for the straightforward detection of various analytes by the naked eye.⁸ In this regard, artificial enzymes or nanozyme-based biosensors have emerged as promising sensors for detecting various analytes due to their improved sensitivity, stability, and cost-effectiveness.⁹

Organic–inorganic hybrid nanozymes show superior catalytic activity and stability compared to single-component nanomaterials.¹⁰ In this context, benzene-containing organic moieties can undergo structural changes upon interaction with target analytes, contributing to enhanced performance. This property can be leveraged in sensor design to detect and quantify target molecules, thereby improving analytical performance in terms of selectivity, sensitivity, and reliability.^11,12 In this context, the study presents novel and easily synthesized nanozymes developed by coordinating copper ions (Cu⁺/Cu²⁺) with gallic acid (GA), a natural polyhydroxy phenolic compound. This organic acid, naturally found in many plants such as hops, sumac, gallnuts, witch hazel, grapes, tea leaves, and oak bark, acts as a potent chelating agent, creating stable complexes with various metals.¹³

In recent years, Cu–GA based nanomaterials have been extensively explored for diverse biomedical and environmental applications owing to their redox activity, coordination versatility, and intrinsic enzyme-mimicking properties. Various Cu–GA systems have been reported, including chitosan–Cu–GA antibacterial nanocomposites with oxidase and peroxidase-like activity for wound healing,¹⁴ aqueously synthesized Cu–GA MOFs as green adsorbents for dye removal,¹⁵ and nanoscale Cu–GA MOFs serving as dual carriers for GA and methylene blue in combined drug delivery and photodynamic therapy.¹⁶ Other studies have integrated Cu–GA MOFs into gentamicin-loaded hydrogels¹⁷ and pomelo peel-based wound dressings¹⁸ to enhance antibacterial and regenerative performance. Ultrasmall Cu–GA nanodots have also been investigated for chemodynamic cancer therapy due to their reactive oxygen species (ROS)-generating capacity in tumor microenvironments.¹⁹ Furthermore, rhodamine-modified Cu–GA nanoparticles have been designed as theranostic agents for Alzheimer's disease, enabling β-Amyloid protein detection, ROS scavenging, and neuroprotection.²⁰ Most recently, a laccase-like Cu–GA MOF with catalytic activity toward substrates such as norepinephrine has also been reported, highlighting its potential for applications in biosensing and environmental remediation.²¹ Despite recent advances, most reported Cu–GA systems are limited to crystalline MOFs or polymer-based composites, primarily developed for antibacterial, therapeutic, or pollutant-remediation applications. However, no reports have been on developing amorphous copper@GA hybrid structures (Cu@GA·HSs) for the selective and sensitive colorimetric detection of SBM, a β-agonist of significant clinical and regulatory relevance. Accordingly, this study aimed to fabricate and characterize Cu@GA·HSs. It also provided comprehensive analytical validation following ICH guidelines and applied the method to real serum samples, with results compared to HPLC. Additionally, molecular docking and dynamics simulations were performed to elucidate the interaction mechanism between SBM and HSs. This combination of analytical application, real-sample validation, and mechanistic insight distinguishes the work from existing Cu–GA studies and highlights its novelty in pharmaceutical analysis.

2. Materials and methods

2.1. Materials

Copper(II) sulfate pentahydrate (CuSO₄·5H₂O), GA, 2,4-dichlorophenol (2,4-DP), and 4-aminoantipyrine (4-AP) were acquired from Sigma-Aldrich (St. Louis, MO, USA). HPLC-grade methanol and acetonitrile were sourced from Merck (Darmstadt, Germany). The Center for Quality Control of Drugs in Tehran, Iran, kindly provided SBM. All other chemicals and reagents used in the experiments were of analytical grade and utilized without additional purification.

2.2. Fabrication and characterization of copper@GA inorganic–organic hybrid structures

Cu@GA·HSs were prepared following a modified literature procedure.²¹ GA (0.2–1 mM) and CuSO₄·5H₂O (2–10 mM) were dissolved in 100 mL of water. The pH of the solution was then adjusted to 7.0 using an aqueous solution of 4 mM sodium hydroxide (NaOH). The reaction mixture was then placed in a sealed tube and heated vigorously at 70 °C. After a 5 h reaction period, the products were isolated by centrifugation at 8000 g and washed three times with distilled water. The resulting precipitate was assayed, and the concentrations of GA and CuSO₄ corresponding to the highest oxidation mimic activity were chosen to prepare HSs for subsequent experiments. In this study, a Tescan MIRA II scanning electron microscope (SEM, Czech Republic) was used to characterize the morphology of the synthesized Cu@GA·HSs. The elemental composition and spatial distribution within the HSs were investigated using energy-dispersive X-ray spectroscopy (EDX) coupled with the same SEM instrument. The Fourier transform infrared (FTIR) spectrum of the synthesized nanozymes was recorded utilizing a Shimadzu Equinox 55 instrument from Japan. X-ray diffraction (XRD) analysis was performed using a PW1730 diffractometer (Philips, Netherlands) to identify the crystalline phases in the synthesized HSs. The thermal stability of the prepared HSs was assessed using a thermogravimetric analyzer (TGA4000, PerkinElmer, USA) in a nitrogen atmosphere. The TGA experiment involved heating the samples from 50 °C to 800 °C at 10 °C min⁻¹ and monitoring the weight changes. The main physicochemical properties and redox activity of the novel Cu@GA·HSs were also characterized using X-ray photoelectron spectroscopy (XPS, Specs-FlexPS, Germany). To evaluate the release profile of Cu²⁺ from the prepared HSs, the structures were immersed in deionized water at 25 °C for 72 h. After incubation, the samples were centrifuged (8000g, 5 min), and the supernatant solutions were analyzed using inductively coupled plasma optical emission spectrometry (ICP-OES) with a SPECTRO ARCOS instrument (Spectro Analytical Instruments, Germany).

2.3. Measuring the oxidase-mimetic activity of Cu@GA·HSs

The oxidase-like activity of Cu@GA·HSs was investigated using 4-AP and 2,4-DP as chromogenic substrates. In each microtube, the reaction mixture (final volume 1.0 mL) consisted of 100 μL of 4-AP solution (1 mg mL⁻¹), 100 μL of 2,4-DP solution (1 mg mL⁻¹), 800 μL of Britton–Robinson (BR) buffer (pH 3.0–8.5, 0.5 intervals) containing Cu@GA·HSs (5 mg mL⁻¹). The mixtures were incubated at different temperatures (20–60 °C, 10 °C intervals) for 20 min. After incubation, the tubes were centrifuged at 10 000g for 10 min to remove HSs. Then, 250 μL of the supernatant was transferred to a 96-well microplate, and the absorbance was measured at 540 nm using a microplate reader (BioTek® Synergy HTX).

2.4. Kinetic studies

The steady-state kinetic analysis of Cu@GA·HSs was performed to evaluate their oxidase-like catalytic activity. The assays were conducted in BR buffer (pH 7.0) at 50 °C using 5 mg mL⁻¹ Cu@GA·HSs, with 4-AP (1 mg mL⁻¹) as a chromogenic mediator and 2,4-DP at varying concentrations (0.25–2 mg mL⁻¹) as the substrate. The absorbance of the oxidized substrate was monitored at 540 nm using absorption spectrophotometry. For comparison, the oxidation reaction was also catalyzed by laccase (0.1 U mL⁻¹) as a natural oxidase, and the results were evaluated alongside those of Cu@GA·HSs. The laccase used in this study was laccase from Trametes versicolor (EC 1.10.3.2), purchased from Sigma-Aldrich (≥0.5 U mg⁻¹). The enzyme was used as received, without further purification. Kinetic parameters were calculated by fitting the data to the Michaelis–Menten eqn (1), where V is the initial velocity, V_max is the maximum reaction rate, C is the substrate concentration, and K_m is the Michaelis constant.

V = V_maxC/(K_m + C) (1)

The kinetic parameters (K_m and V_max) were determined by plotting the initial reaction velocity (V₀) against substrate concentration ([S]) and performing non-linear regression fitting to the Michaelis–Menten model. The V₀ value was obtained from the linear slope of the absorbance–time curve (ΔA/Δt) at 540 nm and converted to concentration units using the Beer–Lambert law (ε = 2.65 × 10⁴ L mol⁻¹ cm⁻¹). To ensure data reliability, all kinetic measurements were performed in triplicate (n = 3), and mean values with corresponding standard deviations were used for data fitting.

2.5. Evaluation of stability and reusability of Cu@GA·HSs

To evaluate the short-term stability of the prepared HSs, they were incubated for 3 h at temperatures ranging from 20 °C to 60 °C in 10 °C intervals and at pH values between 3.0 and 8.5 in 0.5 unit intervals. Following incubation, the samples were centrifuged at 8000g for 5 min. The catalytic activity of HSs was then measured under the selected pH and temperature assay conditions (refer to section 2.3 for details). The results were expressed as the percentage of activity retained after incubation relative to the initial activity (considered 100%).

The HSs were stored at 4 °C, 25 °C, and 37 °C for storage stability analysis. Their catalytic activity was measured every 5 days over 50 days. The results were presented as the percentage of residual activity compared to the initial estimated value. To assess the reusability of the Cu@GA·HSs, the catalysts were recovered by centrifugation after each reaction cycle and washed with water for subsequent reaction cycles.²²

2.6. Analytical application of Cu@GA·HSs for SBM detection

2.6.1. Validation study. The method was validated according to the ICH guidelines for ligand-binding assays (LBA).^23,24 The validation assessed parameters such as linearity, recovery, intra- and inter-day precision, accuracy, limit of quantification (LOQ), and limit of detection (LOD). In brief, linearity was assessed by plotting the sensor response against the concentration of SBM. Method accuracy was demonstrated by assaying three different SBM concentrations (200, 1500, and 3500 μg mL⁻¹) in triplicate. Repeatability and intermediate precision were evaluated by measuring these same concentrations (200, 1500, and 3500 μg mL⁻¹) six times on the same day and over three consecutive days, respectively. The LOD, LOQ, and selectivity were determined following IUPAC guidelines.

2.6.2. Analysis of SBM in serum samples. Sheep serum was purchased from HuraTeb Pharmed (Tehran, Iran). According to the manufacturer, the serum was collected aseptically from healthy sheep with adherence to ethical guidelines, pooled, and sterile-filtered. Red blood cells were removed using appropriate procedures to avoid hemolysis. Prior to analysis, serum samples were centrifuged at 8000g for 30 min at room temperature, and the supernatant was collected and stored at −20 °C. To precipitate serum proteins, 1.5 mL acetonitrile was added to the supernatant, followed by sonication for 1 min in an ultrasonic bath and centrifugation at 10 000g for 10 min. The resulting supernatant was filtered through a 0.45 μm Millipore® filter to obtain protein-free serum. The filtrate was spiked with SBM at final concentrations of 125–5000 μg mL⁻¹. For the assay, 900 μL of spiked serum was mixed with 100 μL of 4-AP solution (1 mg mL⁻¹) in a microtube and incubated under shaking conditions at 50 °C for 20 min. After centrifugation (10 000g, 10 min), 250 μL of the supernatant was transferred to a 96-well plate, and absorbance was measured at 540 nm using the microplate reader.

Eqn (2) was employed for each sample to determine the accuracy of detecting SBM in the serum matrix.

Accuracy of detection (%) = (Measured SBM/Spiked SBM) × 100 (2)

2.6.3. Interference studies. The selective capability of the fabricated sensor was investigated by detecting 1 mM, 3 mM, and 6 mM of several potential co-existing electroactive species, including catechol, levofloxacin, ascorbic acid, ciprofloxacin, glucose, sucrose, fructose, galactose, boric acid, and starch in protein-free serum spiked with SBM (200 μg mL⁻¹). The effect of metal ions on the activity of the prepared nanozyme was also evaluated by adding 1 mM, 5 mM, and 10 mM of various interfering cations, including sodium chloride (NaCl), potassium chloride (KCl), calcium chloride (CaCl₂), lithium chloride (LiCl), copper sulfate (CuSO₄), barium nitrate (Ba(NO₃)₂), zinc chloride (ZnCl₂), iron(II) chloride (FeCl₂), bismuth nitrate (Bi(NO₃)₃), magnesium sulfate (MgSO₄), nickel(II) chloride (NiCl₂), lead(II) acetate (Pb(C₂H₃O₂)₂), cadmium nitrate (Cd(NO₃)₂), chrome nitrate (Cr(NO₃)₃), aluminum chloride (AlCl₃), and iron(III) chloride (FeCl₃) to the serum samples. Additionally, the catalytic performance of the synthesized HSs was evaluated in the presence of 20%, 40%, and 60% (v/v) of organic solvents spanning a wide range of log P values (i.e., methanol (log P −0.69), tetrahydrofuran (log P −0.59), acetonitrile (log P −0.33), acetone (log P −0.24), ethanol (log P −0.24), propanol (log P 0.32), ethylacetate (log P 0.7), 1-butanol (log P 0.8), diethyl ether (log P 0.98), dichloromethane (log P 1.25), isoamyl alcohol (log P 1.3), chloroform (log P 1.97), benzene (log P 2.0), toluene (log P 2.5), petroleum benzene (log P 2.8), 1-octanol (log P 3.0), p-xylene (log P 3.1), cyclohexane (log P 3.6), heptane (log P 4.2), decanol (log P 4.75), n-nonane (log P 5.3), n-dodecane (log P 6.8), n-tetradecane (log P 7.6), n-hexadecane (log P 8.8)). Additionally, 10 mM, 20 mM, and 30 mM of surfactants, including Tween 20, Tween 40, Tween 80, Tween 85, cetyltrimethylammonium bromide (CTAB), Triton X, and sodium dodecyl sulfate (SDS), were introduced into plasma containing SBM to assess the sensor's selectivity.

The selectivity coefficient in the presence of each interfering substance was then calculated using eqn (3), where C₀ represents the initial concentration of spiked SBM, and C_e is the detected concentration of the drug in plasma. A selection coefficient closer to one indicated enhanced selectivity for SBM compared to the interfering substance. All experiments were conducted in triplicate over three days.

Selectivity coefficient (%) = C_e/C₀ (3)

2.6.4. Comparison of HSs with the HPLC method for the determination of SBM. To evaluate the analytical performance of the developed sensor, SBM determination results were compared with those obtained by HPLC. In this regard, a calibration curve for SBM detection was constructed using HPLC. Standard solutions ranged from 125 μg mL⁻¹ to 2000 μg mL⁻¹, and the peak area was used for quantification with triplicate measurements at each concentration. The HPLC separation was achieved using a C18 reversed-phase column (250 × 4.6 mm, 10 μm particle size) with an isocratic mobile phase composed of 10 mM potassium phosphate buffer (pH 3) and methanol (55 : 45, v/v) at a flow rate of 1 mL min⁻¹. A Smartline Autosampler 3950, equipped with a 100 μL sample loop, was used for sample injection. Data acquisition and analysis were performed using ChromGate version 3.3.1 (Knauer, Berlin, Germany).

2.7. Molecular simulation of the sensor preparation and salbutamol detection mechanism

To gain molecular-level insights into the interaction mechanism between SBM and the sensor, computational docking simulations were performed using Materials Studio (BIOVIA, v2023). The polymeric crystal structure of the sensor was constructed based on its monomeric unit using Material Studio software. As illustrated in Fig. 1, the simulated polymer adopts an orderly, periodic configuration, forming a highly symmetrical crystal lattice. The structure comprises regularly repeating monomer units highlighted in the magnified views. These views reveal the detailed bonding interactions and the spatial arrangement of atoms within the polymer chain. Notably, the polymer exhibits a distinct hexagonal morphology, suggesting long-range order and structural uniformity, which may play a critical role in determining its physical and functional properties.

Fig. 1 Multi-scale structural representation of the polymer crystal. (a) Periodic crystalline arrangement of polymer chains, modeled using Material Studio. (b) Ordered packing of polymer chains showing a regular, symmetric pattern. (c) Close-up view of inter-chain interactions and repeating unit connectivity. (d) Chemical structure of the monomer, highlighting aromatic rings and the central coordinating atom that dictates polymer geometry and crystalline order. Atom color scheme: white = hydrogen, green = carbon, red = oxygen, chocolate brown = copper.

The interaction between SBM and the polymer matrix was investigated using the Adsorption Locator module in Materials Studio. The 3D structure of SBM was constructed and geometry-optimized with the Forcite module employing the COMPASS III force field to ensure a stable ligand conformation. The polymer framework was derived from the periodic crystal model and partially fixed to maintain structural integrity during docking, while the binding cavity was defined by selecting target atoms around the central pore.

Docking simulations were performed using a simulated annealing/Monte Carlo protocol, which systematically explores ligand translations, rotations, and regrowth within the defined region. A cutoff distance of 12.5 Å was used for both van der Waals and electrostatic interactions, with cubic spline truncation. Each run consisted of 100 000 loading steps and 15 heating–cooling cycles to ensure sufficient conformational sampling. The convergence of the geometry optimization is illustrated in Fig. 2a, showing stabilization of energy and gradient values over successive steps.

Fig. 2 (a) Preparation of the polymer structure for molecular simulations. (a) Geometry optimization convergence of the modified polymer structure. (b) Positioning of the molecular docking grid box over the central binding region of the modified polymer.

To rationalize the experimental selectivity, docking studies were performed with common serum interferents and doping-related drugs (adrenaline, creatinine, uric acid, glucose, and salmeterol). Comparisons of docking energies and interaction patterns with SBM enabled the identification of the molecular interactions responsible for the sensor's high selectivity. The interaction (binding) energy for each ligand–receptor complex was calculated according to the standard definition:

ΔE = E_(complex) − E_(receptor) − E_(ligand) (4)

Where E_(complex), E_(receptor), and E_(ligand) are the total energies of the optimized complex, the isolated receptor, and the isolated ligand, respectively.

3. Results and discussion

3.1. Design and structural characterization of Cu@GA·HSs

The colorimetric sensor was synthesized by Cu²⁺-mediated oxidative coupling assembly of GA. Due to its strong chelating ability, the self-oxidation of this organic acid was greatly accelerated in the presence of Cu²⁺. The catechol group in GA was converted into reactive semi-quinone and quinone species, which then participated in self-polymerization and formed complexes with Cu²⁺ ions.²⁵ The concentrations of CuSO₄ and GA were two critical factors affecting the activity of the prepared HSs. Copper exhibits exceptional catalytic efficiency even at low concentrations, owing to its robust redox activity and structural stability.²⁶ In heterogeneous catalysis, copper-based systems demonstrate significantly accelerated reaction kinetics compared to alternative catalysts. However, determining the optimal catalyst dose is essential, as an excess amount can lead to scavenging effects, ultimately decreasing process efficiency. Therefore, nanozymes were synthesized by varying GA concentrations from 0.2 mM to 1 mM and CuSO₄ concentrations from 2 mM to 10 mM. Among the resulting HSs, those prepared with 0.6 mM GA and 8 mM CuSO₄ exhibited the most pronounced oxidative activity and were selected for further experimentation (Fig. 3).

Fig. 3 Effect of copper(II) sulfate (CuSO₄) and gallic acid concentrations on (a) oxidase activity and (b) corresponding absorption spectra of the prepared copper@gallic acid hybrid structures (Cu@GA·HSs).

SEM analysis of GA revealed its unique morphology, which is characterized by irregularly shaped crystalline structures. The crystal surfaces appeared smooth, and their size distribution varied significantly. Following the preparation of Cu@GA·HSs, the nanozymes exhibited a rough, irregular morphology and increased specific surface area. Moreover, the particle size grew noticeably during the formation of the HSs (Fig. 4).

Fig. 4 Scanning electron microscope (SEM) images and energy-dispersive X-ray (EDX) mapping of gallic acid and copper@gallic acid hybrid structures (Cu@GA·HSs). SEM images show the morphology of gallic acid (a) and Cu@GA·HSs (e). EDX mapping reveals the distribution of oxygen (b and f), carbon (c and g), and copper (d and h) within these structures.

The detection of copper using EDX analysis following the preparation of HSs indicated its potential incorporation into the GA framework (Fig. 4).

The structural characteristics of GA and the synthesized HSs were further analyzed using FTIR and XRD, with full spectra provided in the SI (Fig. S1 and S2). Briefly, the FTIR spectra indicated the presence of hydroxyl and carbonyl groups in GA, and changes in peak positions and intensities in HSs suggested the formation of metal–ligand bonds. XRD analysis showed that GA exhibited a crystalline nature, while the HSs displayed reduced crystallinity, consistent with the incorporation of GA into the nanozyme matrix.

TGA was employed to characterize the thermal stability and physicochemical properties of the prepared HSs by measuring their weight loss as a function of temperature. The initial stage of GA thermal decomposition involves the gradual evaporation of moisture, concurrent with water produced by self-condensation reactions (100–150 °C). Subsequently, a second decomposition step occurs at (250–400 °C), an optimal temperature for converting phenol into pyrocatechol and modifying short substituents on the benzene ring.²⁷ The curved shape of the thermogram changed after the preparation of Cu@GA·HSs (Fig. S3). The initial weight loss observed at 70 °C was attributed to removing free and chemically bound adsorbed water. The onset of dehydroxylation of GA in the prepared HS was detected at 300 °C, which was higher than the decomposition temperature of the pure sample. XPS analysis of the synthesized HSs provided detailed insights into their surface composition and chemical states, which were crucial for understanding their oxidase-like activity (Fig. 5).

Fig. 5 (a) X-ray photoelectron spectroscopy (XPS) spectrum of the as-synthesized Cu@GA·HSs; (b and c) high-resolution signals corresponding to Cu(I) and Cu(II); and (d and e) deconvoluted peaks assigned to metal-bound oxygen, phenolic/carboxylic oxygen, and adsorbed water or hydroxyl groups, confirming the formation of the Cu–gallate coordination complex.

The Cu 2p₃/₂ peak, observed around 536–559 eV indicated the presence of Cu(I) and Cu(II), with the absence of pronounced satellite peaks suggesting a predominance of Cu(I). The C 1s spectrum (∼1193–1211 eV) revealed components corresponding to aromatic/phenolic carbons (C–C/C–H at ∼284.6 eV), hydroxyl/carboxyl groups (C–O at ∼286–287 eV), and carboxylate/ester functionalities (O–C=O at ∼288–289 eV), confirming partial retention of GA's organic structure and possible coordination with copper ions. The broad O 1s peak (∼947–964 eV) encompassed signals from metal-bound oxygen (Cu–O at ∼530–531 eV), phenolic/carboxylic oxygen (C–O at ∼532–533 eV), and adsorbed water or hydroxyl groups (∼534 eV), supporting the formation of a Cu–gallate coordination complex. The coexistence of redox-active Cu(I) and Cu(II) species within the GA matrix likely drove the observed oxidase-like behavior by enabling efficient electron transfer during substrate oxidation. Additionally, the carbon-rich framework might stabilize copper sites and enhance substrate interactions. The copper ion release from the sensor during a 72 h incubation was only 7 ppm. It indicates that the nanozymes were highly stable, and the metal ions were released slowly from the sensing membrane.

3.2. The effect of the pH and temperature on the activity and stability of the prepared HSs

As previously described (section 2.3), the catalytic activity of Cu@GA·HSs was evaluated using a chromogenic reaction. In the presence of HSs, 2,4-DP was the substrate, reacting with 4-AP to generate the colored product. While neither 2,4-DP nor 4-AP absorbed light at 520 nm, their reaction product exhibited a distinct absorption at this wavelength. The catalytic activity of the prepared HSs was then evaluated at various pH values ranging from 3.0 to 8.5 (Fig. 6a). The maximal activity was observed at pH 7, with a noticeable decline in performance under both acidic (pH 3.0) and alkaline (pH 8.5) conditions. Furthermore, the nanozyme's activity exhibited a temperature-dependent profile. As the temperature increased from 20 °C to 50 °C, catalytic activity steadily increased, reaching a maximum at 50 °C. However, further elevation to 60 °C resulted in a decrease in activity.

Fig. 6 (a) The oxidase-activity and (b) short-term stability of the prepared copper@gallic acid hybrid structures (Cu@GA·HSs) (5 mg mL⁻¹) after incubation under varying pH conditions (3.0–8.5, pH 0.5 intervals) and temperatures (20–50 °C, 10 °C intervals). (c) The reusability of the prepared Cu@gallic acid·HSs (5 mg mL⁻¹) for the reaction between 4-AP and 2,4-DP in 100 mM BR buffer (pH 7) at 50 °C. (d) Storage stability of the prepared HSs over 50 days of incubation at 4 °C, 25 °C, and 37 °C. The error bars show the standard deviation (SD) from three independent measurements (n = 3).

The stability of the catalyst was evaluated under various pH and temperature conditions, as illustrated in Fig. 6b. The results demonstrate a positive correlation between pH and nanonzymes stability within the range of 3 to 7. However, deviations from the natural pH (pH 7) to alkaline conditions (pH 7.5 to 8.5) resulted in a decrease in stability, although less pronounced than in acidic conditions (pH 3, 3.5, and 4). The prepared nanozymes exhibited superior stability at 20 °C and 30 °C compared to higher temperatures (50 °C and 60 °C). The synthesized Cu@GA·HSs function similarly to natural oxidase enzymes, where Cu⁺ oxidizes the phenolic substrate, transferring electrons to Cu²⁺, and reducing molecular oxygen to water. Literature highlights a key advantage of copper-based heterogeneous catalysts: their ability to perform well across a broad pH range, especially near-neutral pH. However, optimal pH depends on the catalyst's characteristics and its point of zero charge. Acidic conditions can decrease catalytic activity and accelerate metal leaching from the catalyst surface. Reaction temperature also significantly influences copper-based catalyst activity. Higher temperatures generally enhance oxidation rates and efficiency, supplying the necessary activation energy.^28,29 However, excessive temperatures may hinder efficiency by forming stable, undesirable oxidation by-products, causing active site loss due to hydration effects, or decomposing the oxidant into inactive species.²⁶

3.3. Storage stability and reusability of the prepared nanozyme

The reusability of the synthesized nanozyme was also assessed through repeated consecutive activity cycles. The experiments were conducted at 37 °C and pH 7.0, using 5 mg L⁻¹ of the catalyst. From Fig. 6c, the catalytic activity decreased by about 30% during the first five recycling cycles, while it rapidly dropped to 60% in the subsequent five cycles. The reduction in catalytic efficiency with increasing cycle times could be attributed to several factors. First, the adsorption of 4-DP molecules onto the nanozyme's active sites (phenolic groups in gallic acid) might hinder substrate access during continuous operation. Repeated washing and separation processes also led to the leaching of Cu ions and the hydrolysis of gallic acid from the prepared HSs. These combined effects contributed to a gradual decline in catalytic activity over time.³⁰ However, the prepared HSs demonstrated superior reusability compared to certain other copper-based catalysts.³¹ For instance, cysteine–Cu catalysts retained less than 60% of their initial activity after five cycles, and nanozymes made of copper and tannic acid lost more than 60% of their activity after the removal of malachite green.³²

The storage stability of nanozymes is crucial for their practical application. The long-term stability of Cu@gallic acid·HSs was evaluated by storing them at 4 °C, 25 °C, and 37 °C for 50 days. As shown in Fig. 6d, the prepared nanozymes retained 60% of their catalytic activity after being stored at 4 °C for 30 days. Moreover, after 40 days of storage at room temperature, HSs retained 41% of their initial catalytic activity. Following storage at 37 °C, the oxidation activity of the catalyst gradually declined, leading to a 70% reduction by the 40th day.

3.4. Kinetic behavior and proposed mechanism of the designed nanozymes

The apparent enzyme kinetic parameters (K_m and V_max) were determined by fitting the experimental data to the Michaelis–Menten equation (Fig. 7a). The fitted curve exhibited an excellent correlation coefficient (R² = 0.999), confirming the reliability of the kinetic analysis. The Cu@GA·HSs nanozyme showed slightly higher substrate affinity (K_m = 0.5 mg mL⁻¹vs. 0.6 mg mL⁻¹) and nearly double the maximum catalytic rate (V_max = 27.9 vs. 14.5 μmol min⁻¹) compared to laccase, resulting in more than twofold greater catalytic efficiency. These findings demonstrate that the nanozyme outperforms the natural enzyme in catalyzing the oxidation of 2,4-DP under identical experimental conditions. Additional experiments were performed to further elucidate the catalytic mechanism of Cu@GA·HSs. Reactive oxygen species (ROS) scavenging experiments were then conducted to identify the active species involved in catalysis. Various scavengers were used: thiourea (for ·OH), DMSO (for O₂·⁻), and sodium sulfite (for O₂). The results indicated that minor amounts of O₂· and ·OH were generated in the reaction system, suggesting they do not significantly contribute to substrate oxidation. Instead, dissolved O₂ was confirmed as the main active species. In particular, sodium sulfite, which removes dissolved oxygen from the solution, caused a dramatic decrease in catalytic activity, completely suppressing the oxidation of substrate to its colored product. Thus, the results clearly demonstrate that Cu@GA·HSs act as oxidase nanozymes, with molecular oxygen serving as the dominant electron acceptor in the catalytic mechanism (Fig. 7b).

Fig. 7 (a) Michaelis–Menten plots for the oxidase-like catalytic activity of Cu@GA·HSs and laccase. (b) Absorption spectra of the prepared Cu@GA·HSs (purple circles) and in the presence of DMSO (orange circles), thiourea (green triangles), and sodium sulfite (navy blue triangles) over the range of 450–550 nm, demonstrating oxidase-like activity.

3.5. Detection of SBM with Cu@GA·HSs

3.5.1. Method validation. The linearity of the proposed spectrophotometric method for SBM determination was evaluated under experimental conditions (Fig. S4). Following the described procedure, six replicates of standard solutions (at various concentrations) were analyzed using the oxidative coupling reaction with 4-AP. The proposed method demonstrated excellent linearity (R² > 0.996), accuracy (96.8–99.8%), and precision (coefficient of variation (CV) 1.1–2.0%) for SBM determination within the concentration range of 125–5000 μg mL⁻¹. The recovery of SBM was evaluated by comparing the concentrations of the drug spiked in phosphate buffer (pH 7.0) with those detected by the sensor. The recovery rates were within acceptable limits, ranging from 74.3% to 82.5% across three different concentration levels, with a CV of ≤10%. To determine the precision and accuracy of the method, standard solutions at concentrations of 200, 1500, and 3500 μg mL⁻¹ were analyzed in triplicate within a single run for intra-day assessment, and across six sets of measurements for inter-day evaluation. The intra- and inter-day accuracy of the method was very high, with values ranging from 97.0–99.7% and 98.0–99.8%, respectively. To assess the method's reproducibility, data from three different batches of HSs were combined and analyzed for inter-day precision and accuracy. Statistical analysis confirmed that the method's performance remained consistent across all batches. The precision (CV%) values of 1.5%, 2.0%, and 3.2% were obtained at concentrations of 200, 1500, and 3500 μg mL⁻¹, respectively. The proposed spectrophotometric system exhibited significantly lower LOD (167 μg mL⁻¹) and LOQ values (220 μg mL⁻¹) compared to other methods, demonstrating its superior sensitivity for the target analyte (Table 1).

Table 1 Validation data for detection of salbutamol by copper@gallic acid hybrid structures (Cu@GA·HSs)

Parameter	Nominal concentration (μg mL^−1)	Back-calculated concentration (μg mL^−1)	Precision (%)	Accuracy (%)
Calibration curves	125	121 ± 2	1.6	96.8
	250	243 ± 5	2.0	97.2
	500	496 ± 9	1.8	99.2
	1000	991 ± 11	1.1	99.1
	2000	1989 ± 24	1.2	99.5
	3000	2991 ± 36	1.2	99.7
	4000	3987 ± 45	1.1	99.6
	5000	4994 ± 67	1.3	99.8
Intra-day	200	194 ± 3	1.5	97.0
	1500	1491 ± 22	1.4	99.4
	3500	3489 ± 48	1.4	99.7
Inter-day	200	196 ± 7	3.5	98.0
	1500	1490 ± 32	2.1	99.3
	3500	3493 ± 93	2.6	99.8
		Recovery (%)	CV (%)
Recovery	200	74.3 ± 1.1	1.5
	1500	78.4 ± 1.6	2.0
	3500	82.5 ± 2.7	3.2

---

3.5.2. A comparison of HSs and HPLC methods for the determination of SBM. The potential of Cu@GA·HSs for SBM assay was evaluated by re-analyzing standard samples that were previously quantified using the standard HPLC reference technique. The correlation between the sensor and the HPLC method is shown in Fig. S5. The slope of the regression line was 1.09, and the Pearson correlation coefficient was 0.997, indicating good agreement between the two methods. The mean difference between the values obtained from the two techniques was 0.06%, indicating a deviation from the reference method of less than 20%. Although the sensor's LOD value (167 μg mL⁻¹) was far from the reference HPLC method, its performance remains competitive with other widely used detection techniques. The main advantage of the prepared HSs over chromatographic and electrochemical methodologies is the elimination of organic solvents during analysis, making it a more environmentally friendly and practical choice for SBM detection. This approach is also faster and requires less specialized expertise, enhancing its accessibility and ease of use.

The median plasma SBM concentration (∼10 μg L⁻¹)³³ is below the LOD of the presented method (167 μg mL⁻¹); therefore, the current system is not intended for plasma pharmacokinetic monitoring. Instead, the use of Cu@GA·HSs is aimed at pharmaceutical quality control and anti-doping applications, where the relevant concentrations are substantially higher. Within this context, the method exhibits adequate sensitivity, remaining well below the World Anti-Doping Agency (WADA) daily upper limit of 1600 μg per day for salbutamol.³³

3.5.3. Selectivity of HSs for the determination of SBM in plasma samples. Catalytic reactions mainly occur on the surface of the catalyst. Therefore, the interface interaction between the catalyst and its environment is crucial to determining its catalytic performance.³⁴ Thus, further exploration was conducted on plasma spiked with SBM in the presence of interfering agents, with a focus on the interaction between the catalyst and the plasma interface (Fig. 8). Based on the data, the interfering agents show varied effects on the selectivity of Cu@GA·HSs catalysts for oxidizing 4-AP and SBM. In contrast to ascorbic acid, catechol, and galactose, which have multiple hydroxyl groups, exhibited greater interference, attributed to their higher potential for competitive adsorption on active sites. Larger and complex ring structures (e.g., ciprofloxacin and levofloxacin) significantly reduced selectivity, suggesting strong interactions with the catalyst surface. Simple sugars, including glucose and sucrose, caused moderate interference, possibly due to their structural flexibility and multiple binding sites (Fig. 8a).

Fig. 8 The effect of (a) interfering agents and (b) surfactants on the selectivity of Cu@gallic acid·HSs (5 mg mL⁻¹) catalysts for the oxidation of 4-aminoantipyrine (4-AP) and SBM. A selectivity value closer to 1 indicates higher selectivity, while deviations from 1 suggest lower selectivity.

Additionally, nonionic surfactants (Tween 85, 80, 60, 40) and Triton X-100 showed increasing interference at higher concentrations, reducing selectivity. While Tween 20 initially maintained selectivity, it decreased at elevated levels. Ionic surfactants (SDS, CTAB) exhibited significant interference at high concentrations, with CTAB having the most pronounced effect on selectivity (Fig. 8b).

The selectivity of Cu@GA·HSs for detecting SBM was assessed in the presence of 1 mM, 5 mM, and 10 mM of various metal salts. CaCl₂, FeCl₃, and NaCl maintained selectivity values close to 1 at all concentrations, indicating minimal interference. AlCl₃, KCl, and CuSO₄ showed slight deviations with increasing concentrations, suggesting moderate interference. Pb(C₂H₃O₂)₂ and NiCl₂ exhibited the highest deviations, consistently around 2, signaling significant interference. In contrast, Ba(NO₃)₂, Bi(NO₃)₃, and ZnCl₂ displayed sharp reductions in selectivity at 10 mM, indicating substantial interference at high concentrations (Fig. 9a). Fig. 9b presents the impact of organic solvents on the selectivity of Cu@GA·HSs for SBM detection. Non-polar solvents such as tetradecane, dodecane, and nonane maintained selectivity close to 1, indicating minimal interference. Solvents with higher log P values, like heptane and cyclohexane, exhibited moderate interference. Aromatic solvents (toluene, benzene) and polar solvents (methanol, dichloromethane) caused significant deviations at higher concentrations, suggesting increased interference likely due to competitive binding. This result indicates that both high- and low-polarity solvents can disrupt the sensor's selectivity, particularly at elevated concentrations. The prepared Cu@GA·HSs offered several advantages, including low cost, ease of preparation, excellent stability, and good durability. However, unprotected nanozymes were susceptible to non-specific binding from interfering molecules, leading to blocked active sites, reduced catalytic activity, and compromised detection of the target drug.^35,36 Additionally, these nanozymes could aggregate in high ionic strength samples, further hindering catalytic activity and introducing interference.³⁰ Under these conditions, an effective strategy is to eliminate interfering agents, while another promising approach involves functionalizing the nanozyme with organic ligands.^36,37

Fig. 9 The selectivity of copper@gallic acid hybrid structures (Cu@GA·HSs) for detection of salbutamol in the presence of (a) metal salts (1 mM, 5 mM, and 10 mM) and (b) organic solvents (20%, 40%, and 60%). Selectivity values closer to 1 indicate better performance, while deviations from 1, whether higher or lower, suggest less optimal performance.

To further validate its practical utility, the analytical performance of the Cu@GA·HSs was comprehensively benchmarked against other advanced methods for SBM detection, as summarized in Table 2. Electrochemical and immunosensors demonstrate superior sensitivity, achieving detection limits in the picomolar to nanomolar range, which is crucial for identifying trace levels in anti-doping contexts.^38–44 However, these techniques often require complex electrode modifications, incorporating biological elements (e.g., antibodies), or extensive sample pre-treatment,⁴² thereby increasing cost, operational complexity, and analysis time.

Table 2 Comprehensive comparison of the analytical performance and practical applicability of the proposed Cu@GA·HSs optical sensor with other reported methods for salbutamol detection

Analytical method/sensor type	Sample matrix	Linear range	Limit of detection (LOD)	Selectivity/key advantages	Real sample performance (recovery %)	Ref.
Cu@gallic acid·HSs optical nanozyme	Validated vs. HPLC	125–5000 μg mL^−1	167 μg mL^−1	High resistance to metal ions, solvents (K ≈ 1). Validated by MD simulations & docking. Excellent long-term stability (60% activity after 24 days)	Excellent agreement with HPLC (R^2 = 0.997, deviation <0.06%)	This work
Tb4O7/RGO electrochemical sensor	Human serum	1–710 μM	21 nM	Excellent selectivity vs. ascorbic acid, uric acid, caffeic acid, glucose, ractopamine, clenbuterol	>96.5% (in human serum)	38
MIP/graphene-PEDOT:PSS nanocomposite electrode	Swine meat & feed	1 nM–1.2 μM	100 pM	High selectivity from MIP vs. structural analogues	>96.5%	39
Ionic liquid@N doped porous carbon–Co3O4 electrochemical sensor	Urine	2 nM–2500 μM	0.0014 nM	Excellent selectivity	(Successfully used in urine)	40
Ag–nitrogen co-doped reduced graphene oxide/MIP electrochemical sensor	Human urine & pork samples	0.03–20 μM	7 nM	Excellent selectivity from MIP	98.9–105.3%	41
CdTe quantum dots (fluorescence)	Pig urine	0.0627–0.209 μM	42 nM	High selectivity vs. 9 other veterinary drugs	(Applied to pig urine)	45
Cadmium-based coordination polymer (fluorescence)	Simulated urine	—	122 nM	Excellent anti-interference & recyclable	(Successfully used in simulated urine)	46
Cu^2+-mediated MIP carbon paste	Plasma, urine	1–55 nM	0.6 nM	Requires SPE pre-treatment for real samples	94.1–96.3% (after SPE)	42
Immunosensor (GCE/GNP/RAC)	Pig lean, fat, & liver sample	—	0.1 ng mL^−1 (multiresidue)	Multiresidue for 6 β-agonists	80.5–106% (variable)	43
Polyamide amine–Au/horseradish oxidase immunosensor	Real samples	0.1–1500 ng mL^−1	0.03 ng mL^−1	High specificity from antibody	Successfully used in real samples	44

---

In contrast, the proposed optical nanozyme sensor operates in a higher concentration range (hundreds of μg mL⁻¹), making it well-suited for rapid, semi-quantitative screening of concentrated samples such as powdered formulations, inhaler solutions, or plasma prior to confirmatory LC-MS/MS analysis. This approach provides a complementary, rapid, and cost-effective alternative in anti-doping applications where detecting high concentrations of salbutamol is required.

3.6. Simulated mechanism of polymer-based sensor for SBM detection

ollowing the docking setup, the interaction between SBM and the polymer was evaluated using Material Studio. Multiple docking runs were performed, and the resulting binding poses were ranked based on their binding energy scores. Following geometry optimization of the polymer framework, the adsorption of SBM was first examined using the Adsorption Locator protocol in Materials Studio. The calculated isosteric heat of adsorption for SBM was −18.9 kcal mol⁻¹, the highest among all ligands studied, indicating a very strong interaction with the polymer cavity. Structural analysis of the top-ranked poses revealed that SBM is accommodated deeply within the central pore of the polymer. Given the presence of hydroxyl and amine groups in SBM, the high adsorption energy is most likely due to multiple hydrogen bonds with linker oxygen atoms and π–π stacking between the aromatic ring of SBM and gallic moieties of the polymer.

To assess selectivity, docking simulations were repeated for common serum interferents (adrenaline, creatinine, uric acid, glucose, and salmeterol). The adsorption-energy distributions show a clear left-to-right ordering of peak positions (more negative → more favorable): salmeterol (≈−26 kcal mol⁻¹) < salbutamol (≈−18) < adrenaline (≈−14) < uric acid (≈−12) ≈ glucose (≈−7.3) < creatinine (≈−11). Thus, aside from salmeterol—which is structurally related to SBM—the other interferents bind less favorably than SBM.

Adrenaline exhibits moderate affinity, consistent with transient hydrogen bonding but without the full set of interactions seen for SBM. Glucose shows intermediate adsorption due to multiple hydroxyls, yet steric mismatch limits stable encapsulation. Creatinine and uric acid display the weakest binding, reflecting poor spatial complementarity with the polymer cavity. Salmeterol's distribution is centered at the most negative energies but is broader, suggesting multiple poses and potential steric crowding despite its favorable energetics relative to SBM.

To further elucidate the origin of selectivity, adsorption-energy distribution plots for SBM and the interferents above was generated. These plots provide a statistical view of the docking energy landscape, allowing comparison of both central tendency and dispersion (Fig. 10). SBM shows a narrow, well-defined minimum around −18 kcal mol⁻¹, indicative of a dominant, highly specific binding mode within the polymer cavity. In contrast, adrenaline has broader distributions at higher energies (∼−14 kcal mol⁻¹), and glucose/uric acid/creatinine peak still further to the right (−12 to −11 kcal mol⁻¹). Salmeterol peaks left of SBM but with greater breadth, consistent with favorable but less uniquely defined binding.

Fig. 10 Adsorption energy distribution profiles of salbutamol compared with potential interfering molecules (adrenaline, salmeterol, glucose, creatinine, and uric acid).

Collectively, these profiles indicate that the polymer framework forms a uniquely stable and specific interaction with SBM, while most small-molecule interferents are either sterically hindered or energetically disfavored. The exception of salmeterol (more favorable energy but broader, less specific distribution) is consistent with its close structural similarity to SBM and larger size, which can introduce pose heterogeneity. This analysis provides a coherent computational rationale for the sensor's experimental selectivity toward SBM (Table 3).

Table 3 Calculated adsorption energies of salbutamol and potential interfering molecules

Molecule	RDAE	Remarks on binding
a Rigid docking adsorption energy (RDAE) (kcal mol^−1).
Salmeterol	−26	More favorable than SBM, but a broader distribution suggests multiple poses and reduced specificity
Salbutamol	−18.3	Narrow peak consistent with a dominant, highly specific binding mode
Adrenaline	−14	Moderate affinity; transient H-bonding, suboptimal fit
Glucose	−12.8	Polar contacts present, but geometric complementarity is limited
Uric acid	−12.6	Multiple OH groups enable H-bonding; steric mismatch prevents stable encapsulation
Creatinine	−10	Weakest binding; poor spatial complementarity with the cavity

---

4. Conclusions

This study developed a novel, cost-effective, and environmentally friendly approach for detecting SBM using Cu@GA·HSs. The hybrid sensor demonstrated satisfactory stability, reusability, and selectivity for detecting SBM, and its analytical performance showed good agreement with the standard HPLC method. Furthermore, molecular docking and simulation provided insights into the binding mechanism between the polymeric sensor and SBM, revealing dynamic interactions consistent with observed sensor behavior. Despite its promising features, the sensor system exhibited some limitations. The detection sensitivity, while sufficient for many applications, was still lower than that of chromatographic techniques. Unprotected active sites on the nanozyme surface were also prone to interference from structurally similar molecules, surfactants, or high ionic strength matrices, potentially affecting selectivity and accuracy. Future studies should focus on improving the sensor's performance by functionalizing the nanozyme surface with selective recognition elements to minimize non-specific interactions. Moreover, miniaturization and integration of the sensor into portable platforms could pave the way for real-time, point-of-care diagnostics. Expanding the platform for multiplexed detection or adapting it for other biologically relevant analytes may further broaden its applicability in clinical, pharmaceutical, and environmental monitoring.

Despite the promising results, this study has certain limitations. The experiments were performed mainly in protein-free solutions and spiked sheep serum, which may not fully capture the complexity and variability of human clinical samples. Thus, further validation using real patient plasma is necessary to establish the sensor's clinical applicability. In addition, optimizing strategies to minimize non-specific interactions and aggregation under high ionic strength conditions could further enhance the selectivity and robustness of the sensor in complex biological environments. Extending this platform to detect other β-agonists or structurally related compounds would also expand its utility in clinical, pharmaceutical, and food safety monitoring.

Author contributions

Fatemeh Parad: methodology, formal analysis, writing original draft, and conceptualization. Fahimeh Ghasemi: methodology, formal analysis, and data curation. Parisa Khadiv-Parsi: methodology, formal analysis, and conceptualization. Haider Al Amili: methodology, formal analysis, and conceptualization. Parna Keramati: methodology, formal analysis, and conceptualization. Somayeh Mojtabavi: supervision, conceptualization, methodology, resources, data curation, writing – review & editing, project administration, and funding acquisition. Mohammad Ali Faramarzi: supervision, conceptualization, methodology, resources, data curation, writing – review & editing, project administration.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data that support the findings of this study are available in the supplementary information (SI) of this article.

Supplementary information: SI includes FTIR, XRD, and TGA characterizations confirming Cu@GA·HSs formation, along with calibration and correlation curves validating the salbutamol detection assay against HPLC results. See DOI: https://doi.org/10.1039/d5me00136f.

Acknowledgements

This work was financially supported by grant number 1403-2-151-73607 from Tehran University of Medical Sciences, Tehran, Iran, to S. M.

References

1. M. Hostrup, A. Kalsen, M. Auchenberg, S. Rzeppa, P. Hemmersbach, J. Bangsbo and V. Backer, Drug Test. Anal., 2014, 6, 528–532,  DOI:10.1002/dta.1568.
2. J. A. Heuberger, S. C. van Dijkman and A. F. Cohen, Br. J. Clin. Pharmacol., 2018, 84, 1830–1838,  DOI:10.1111/bcp.13619.
3. P. L. Sahu, A. K. Maurya, S. Priyadarshini, A. A. Siddiqui and K. Sahu, J. Chromatogr. Sci., 2023, 61, 605–611,  DOI:10.1093/chromsci/bmac079.
4. Y. Zhang, Alexandria Eng. J., 2024, 96, 185–194,  DOI:10.1016/j.aej.2024.03.087.
5. P. Courlet, T. Buclin, J. Biollaz, I. Mazzoni, O. Rabin and M. Guidi, CPT: Pharmacometrics Syst. Pharmacol., 2022, 11, 469–481,  DOI:10.1002/psp4.12773.
6. P. Raksawong, K. Chullasat, P. Nurerk, P. Kanatharana, F. Davis and O. Bunkoed, Anal. Bioanal. Chem., 2017, 409, 4697–4707,  DOI:10.1007/s00216-017-0466-8.
7. R. Chokkareddy, J. C. Kilele, S. Kanchi and G. G. Redhi, Curr. Anal. Chem., 2023, 19, 220–239,  DOI:10.2174/1573411019666230117152951.
8. K. Davoodi-Rad, A. Shokrollahi, F. Shahdost-Fard, K. Azadkish and E. Madani-Nejad, Microchim. Acta, 2024, 191, 168,  DOI:10.1007/s00604-024-06238-y.
9. E. Madani-Nejad, A. Shokrollahi and F. Shahdost-Fard, Spectrochim. Acta, Part A, 2023, 296, 122665,  DOI:10.1016/j.saa.2023.122665.
10. K. Vojdanitalab, H. Jafari-Nodoushan, S. Mojtabavi, M. Shokri, H. Jahandar and M. A. Faramarzi, Sci. Rep., 2022, 12, 9297,  DOI:10.1038/s41598-022-13490-w.
11. S. Jain, B. Sharma, N. Thakur, S. Mishra and T. K. Sarma, ACS Appl. Nano Mater., 2020, 3, 7917–7929,  DOI:10.1021/acsanm.0c01415.
12. X. Ren, J. Liu, J. Ren, F. Tang and X. Meng, Nanoscale, 2015, 7, 19641–19646,  10.1039/C5NR04685H.
13. B. Badhani, N. Sharma and R. Kakkar, RSC Adv., 2015, 5, 27540–27557,  10.1039/C5RA01911G.
14. X. Sun, M. Dong, Z. Guo, H. Zhang, J. Wang, P. Jia, T. Bu, Y. Liu, L. Li and L. Wang, Int. J. Biol. Macromol., 2021, 167, 10,  DOI:10.1016/j.ijbiomac.2020.11.153.
15. B. Azhar, A. E. Angkawijaya, S. P. Santoso, C. Gunarto, A. Ayucitra, A. W. Go, P. L. Tran-Nguyen, S. Ismadji and Y. H. Ju, Sci. Rep., 2020, 10, 19212,  DOI:10.1038/s41598-020-75927-4.
16. S. Sharma, D. Mittal, A. K. Verma and I. Roy, ACS Appl. Bio Mater., 2019, 2, 2092,  DOI:10.1021/acsabm.9b00116.
17. J. Yang, Z. Huang, J. Tan, J. Pan, S. Chen and W. Wan, Bioact. Mater., 2024, 32, 260,  DOI:10.1016/j.bioactmat.2023.10.005.
18. H. Hamedi, S. Javanbakht and R. Mohammadi, J. Ind. Eng. Chem., 2024, 133, 454,  DOI:10.1016/j.jiec.2023.12.02.
19. Y. N. Hao, Y. R. Gao, Y. Li, T. Fei, Y. Shu and J. H. Wan, Adv. Mater. Interfaces, 2021, 8, 2101173,  DOI:10.1002/admi.202101173.
20. L. Liu, W. Liu, X. Dong and Y. Sun, Chem. Eng. J., 2025, 163063,  DOI:10.1016/j.cej.2025.163063.
21. O. G. Hussein, Y. Mohamed, N. Mostafa and A. M. Mahmoud, RSC Adv., 2025, 15, 11730,  10.1039/D5RA00942A.
22. A. Ansari, S. Vahedi, O. Tavakoli, M. Khoobi and M. A. Faramarzi, Appl. Organomet. Chem., 2019, 33, e4634,  DOI:10.1002/aoc.4634.
23. ICH Guideline, Bioanalytical method validation and study sample analysis M10, ICH Harmonised Guideline, 2022.
24. A. Puscasu, M. Zanchetta, B. Posocco, D. Bunka, S. Tartaggia and G. Toffoli, Anal. Bioanal. Chem., 2021, 413, 1225–1236,  DOI:10.1007/s00216-020-03087-5.
25. S. Lin, Y. Cheng, H. Zhang, X. Wang, Y. Zhang, Y. Zhang, L. Miao and X. Zhao, Small, 2020, 16, 1902123,  DOI:10.1002/smll.201902123.
26. S. Hussain, E. Aneggi and D. Goi, Environ. Chem. Lett., 2021, 19, 2405–2424,  DOI:10.1007/s10311-021-01185-z.
27. D. Dorniani, M. Z. Hussein, A. U. Kura, S. Fakurazi, A. H. Shaari and Z. Ahmad, Int. J. Nanomed., 2012, 7, 5745–5756,  DOI:10.2147/IJN.S35746.
28. L. Lyu, L. Zhang and C. Hu, Chem. Eng. J., 2015, 274, 298–306,  DOI:10.1016/j.cej.2015.03.137.
29. M. A. Faramarzi, M. Mogharabi-Manzari and H. Brandl, Hydrometallurgy, 2020, 191, 105228,  DOI:10.1016/j.hydromet.2019.105228.
30. Z. Ge, B. Wu, T. Sun, B. Qiao and L. Qiao, Colloid Polym. Sci., 2021, 299, 1533–1542,  DOI:10.1007/s00396-021-04867-w.
31. Y. Tang, S. Jiang, W. Li, S. J. Shah, Z. Zhao, L. Pan and Z. Zhao, Chem. Eng. J., 2022, 448, 137701,  DOI:10.1016/j.cej.2022.137701.
32. N. Karikalan, R. Karthik, S. M. Chen, M. Velmurugan and C. Karuppiah, J. Colloid Interface Sci., 2016, 483, 109–117,  DOI:10.1016/j.jcis.2016.08.028.
33. F. Pillard, M. Lavit, V. L. Cances, J. Rami, G. Houin, A. Didier and D. Rivière, Respir. Res., 2015, 16, 155,  DOI:10.1186/s12931-015-0315-2.
34. S. Mojtabavi, M. R. Khoshayand, M. R. Fazeli, N. Samadi and M. A. Faramarzi, Environ. Technol. Innovation, 2020, 20, 101046,  DOI:10.1016/j.eti.2020.101046.
35. C. F. Peng, Y. Y. Zhang, L. Y. Wang, Z. Y. Jin and G. Shao, Anal. Methods, 2017, 9, 4363–4370,  10.1039/C7AY01317E.
36. B. Unnikrishnan, C. W. Lien, H. W. Chu and C. C. Huang, J. Hazard. Mater., 2021, 401, 123397,  DOI:10.1016/j.jhazmat.2020.123397.
37. F. Peytam, F. S. Hosseini, M. Hekmati, B. Bayati, M. S. Moghadam, Z. Emamgholipour, L. Firoozpour, S. Mojtabavi, M. A. Faramarzi, S. E. Sadat-Ebrahimi and M. B. Tehrani, Sci. Rep., 2023, 13, 15672,  DOI:10.1038/s41598-023-42549-5.
38. Y. Zhang, Alexandria Eng. J., 2024, 96, 185,  DOI:10.1016/j.aej.2024.03.087.
39. D. Dechtrirat, B. Sookcharoenpinyo, P. Prajongtat, C. Sriprachuabwong, A. Sanguankiat, A. Tuantranont and S. Hannongbua, RSC Adv., 2018, 8, 206,  10.1039/C7RA09601A.
40. M. U. Rehman, G. A. Mersal, M. F. Farid, A. A. Al-Ghamdi, A. I. Al-Sulami, M. W. Fazal, M. M. Ibrahim, S. Majeed and N. Akhtar, New J. Chem., 2023, 47, 18411,  10.1039/D3NJ03230B.
41. J. Li, Z. Xu, M. Liu, P. Deng, S. Tang, J. Jiang, H. Feng, D. Qian and L. He, Biosens. Bioelectron., 2017, 90, 210,  DOI:10.1016/j.bios.2016.11.016.
42. T. Alizadeh and L. A. Fard, Anal. Chim. Acta, 2013, 769, 100,  DOI:10.1016/j.aca.2013.01.032.
43. C. Gu, P. Ren, F. Zhang, G. Zhao, J. Shen and B. Zhao, ACS Omega, 2020, 5, 5548,  DOI:10.1021/acsomega.0c00249.
44. X. Han, Int. J. Electrochem. Sci., 2020, 15, 7337,  DOI:10.20964/2020.08.93.
45. J. Zhang, Y. Wei, S. Qiu and Y. Xiong, Spectrochim. Acta, Part A, 2021, 247, 119107,  DOI:10.1016/j.saa.2020.119107.
46. Y. Jia, D. Li, N. Wu, S. Su, S. Qie and M. Hu, New J. Chem., 2023, 47, 674,  10.1039/D2NJ05079J.

---