Graphite powder, chloroauric acid (HAuCl₄·4H₂O), and the ACV standard powder were purchased from Aladdin Reagent Co. Ltd. Sodium citrate, NaH₂PO₄, Na₂HPO₄, H₃PO₄, NaOH, and other chemical reagents were purchased from Sinopharm Chemical Reagent Co., Ltd. ACV tablets (200 mg) were purchased from the local pharmacies. All other chemicals and reagents used in this work were of analytical grade and used directly. ACV stock solution (1.0 mM) was prepared by dissolving appropriate amounts of the ACV powder in ultrapure water. Before use, it was stored at 4°C in the dark to avoid any decomposition. The required concentration of ACV was made by diluting the stock solution. The supporting electrolyte was 0.1 M phosphate buffer (PB) solution. The pH varying from 5.5 to 8.0 was obtained by mixing NaH₂PO₄ and Na₂HPO₄, using H₃PO₄ and NaOH as reagents for pH adjusting. Ultrapure water (18 MΩ cm) obtained from a Milli-Q water purifying system was used for preparation of all the solutions.

The structure and surface morphology of the prepared products were observed by transmission electron microscopy (TEM) (JEOL JEM-2100F) operated at 200 kV. The surface composition and chemical state of the rGO–TiO₂-–Au nanocomposites were examined by X-ray photoelectron spectroscopy (XPS) with a monochromatic Al Kα excitation source. The crystal phase of the materials was investigated by using a Bruker D8 Advance X-ray diffractometer at 40 kV with Cu Kα radiation (λ = 1.54 Å). The electrochemical performance of the nanocomposites was measured using a CHI 660E electrochemical workstation. A standard three-electrode system was engaged, which comprised a GC (bare or modified) working electrode with a diameter of 3 mm, a platinum wire counter electrode, and a 3-M KCl-saturated Ag/AgCl reference electrode. All the electrochemical experiments were carried out at room temperature with dissolved oxygen removed by a N₂ stream.

GO, TiO₂, and Au nanoparticles were first prepared. GO was prepared by the oxidation of the graphite powder using a modified Hummer’s method. Experimental details were given in the literature (Kong et al., 2015). TiO₂ nanoparticles were synthesized by the hydrothermal method using tetrabutyl titanate as the titanium source (Ya et al., 2015). Au nanoparticles were synthesized by the citrate reduction method. Details can be found in our previous work (Kong et al., 2019).

For the synthesis of rGO–TiO₂–Au nanocomposites, 10 mg of GO was dispersed in 20 ml ultrapure water and sonicated for 30 min to obtain a uniform dispersion. Then, 5.0 ml of GO dispersion was mixed with 100 µl TiO₂ and sonicated for 1.0 h. After adding 100 µl Au nanoparticles and ultrasonic treatment for another 0.5 h, the mixture was poured into a transparent vial and illuminated using a UV-LED spot lamp. After irradiation for 3 h, the product was collected by centrifugation and washed several times with ultrapure water.

The preparation process of rGO–TiO₂–Au/GCE was as follows: initially, the GCE was carefully polished with 0.3 and 0.05 μm alumina powder in sequence on a polishing cloth to obtain a mirror-like surface. After ultrasonic cleaning in 1:1 nitric acid, ethanol, and ultrapure water for 5 min, it was dried under N₂ flow. Then, 7 μl of rGO–TiO₂–Au dispersion was drop-cast on the surface of GCE using a microsyringe, and it was dried naturally under a closed vessel.

The electrochemical measurements were performed in 10 ml of 0.1 M PB solution containing a certain concentration of ACV using cyclic voltammetry (CV) and linear sweep voltammetry (LSV). The accumulation step was performed at open-circuit for 80 s under stirring. After 5 s quiescence, LSV was recorded between +0.6 and +1.4 V. The oxidation peak currents were measured at 1.10 V for the quantification of ACV.

A total of ten tablets of ACV were finely powdered using the agate and a mortar, and a portion of this powder was accurately weighed and dissolved in ultrapure water with ultrasonic agitation for 10 min to ensure complete dissolution. Finally, it was filtered and diluted to volume with ultrapure water. A desired volume of the sample solution was transferred to the electrochemical cell and analyzed under optimal conditions using the LSV technique. The ACV content in the tablet was calculated using the standard addition method.

The microstructure and morphology of the prepared nanomaterials are observed by TEM, and the obtained images are displayed in Figure 1. Figure 1A depicts the TEM micrograph of GO, in which an ultra-thin, wrinkled, and sheet-like structure is observed. The TEM image of TiO₂ exhibits a small amount of aggregation, which is composed of spheroid nanoparticles with a diameter of about 10 nm (Figure 1B). Figure 1C shows the well-monodispersed and spherical shape of Au nanoparticles with a mean size of about 15 nm. The obtained nanocomposites maintained the 2D sheet structure as shown in Figure 1D. TiO₂ and Au nanoparticles were loaded on the rGO sheet and accumulated along the wrinkles and edges.

XRD analysis was recorded to study crystalline characteristics of the nanocomposites, and the results are presented in Figure 2. As can be observed, GO reveals a sharp and intensive diffraction peak at 2θ of 11.0°, which corresponds to the (002) plane of the graphene sheets (Wang et al., 2008), indicating the formation of GO by the Hummers’ method. The XRD pattern of rGO–TiO₂ shows the diffraction peaks at 25.3°, 37.8°, 48.0°, 55.1°, and 62.7°, which can be indexed to (101), (004), (200), (211), and (204) crystal planes of anatase TiO₂ (JCPDS card no.21-1272) (Li et al., 2007), respectively. Meanwhile, the characteristic diffraction peak of GO disappears, indicating the successful reduction of GO via UV irradiation. According to a previous report (Kong et al., 2016; Yang et al., 2016), when TiO₂ is exposed to UV light, the photo-induced electron–hole pairs are produced. The separated holes react with water to generate oxygen and H⁺, whereas the electrons are effectively captured by the GO substrate to reduce functional groups. The XRD pattern of the rGO–TiO₂–Au nanocomposites is similar to rGO–TiO₂, suggesting that the introduction of Au nanoparticles did not alter their lattices.

The formation of rGO–TiO₂–Au nanocomposites and their surface features were examined by XPS, and the corresponding results are illustrated in Figure 3. The presence of major elements such as Au, C, Ti, and O from the survey spectrum conveys the successful preparation of rGO–TiO₂–Au nanocomposites (Figure 3A). For the Ti 2p spectrum, two main peaks appeared at 459.4 and 464.1 eV fit binding energy of Ti 2p_3/2 and Ti 2p_1/2 (Figure 3B), confirming Ti ions occur in the form of Ti⁴⁺ states. The result is consistent with the reported literature (Zhu et al., 2020). In Figure 3C, Au 4f displays two peaks located at a binding energy of 84.17 and 87.82 eV, which are assigned to Au 4f_7/2 and Ag 4f_5/2 of metallic Au, respectively (Shukla et al., 2018).

The electrochemical oxidation of ACV at the rGO–TiO₂–Au/GCE was investigated by the CV method (Figure 4). In the potential range from +0.6 to +1.4 V, the CV in the absence of ACV shows no observable redox peaks (Figure 4A, curve a). However, in the presence of ACV, a well-resolved oxidation peak is observed at about 1.10 V (Figure 4A, curve b), indicating that the oxidation peak is attributed to ACV. Furthermore, no reduction peaks are found in the reverse scan, suggesting that the electrochemical reaction is a totally irreversible process. Based on the aforementioned results and previously published literatures (Wang et al., 2013; Dorraji and Jalali, 2016; Ilager et al., 2021), a possible oxidation mechanism of ACV involves the two-electron and two-proton transfer process for the formation of 8-oxoacyclovir, which is structurally analogous to the preliminary oxidation product of guanine (Supplementary Figure S1).

The electrochemical behavior of ACV at various electrodes included rGO–TiO₂–Au/GCE (a), rGO–Au/GCE (b), rGO–TiO₂/GCE (c), rGO/GCE (d), TiO₂/GCE (e), and bare GCE (f) as illustrated in Figure 4B. As can be seen, the CV of ACV shows a broad peak and poor current response at bare GCE, revealing sluggish electron-transfer kinetics. At TiO₂/GCE and rGO/GCE, the peak current increases due to the huge surface area of nanomaterials. In great contrast, a well-defined and resolved oxidation peak of ACV can be observed at rGO–TiO₂/GCE, rGO–Au/GCE, and rGO–TiO₂–Au/GCE. Most notably, the rGO–TiO₂–Au/GCE shows the highest augmentation toward the determination of ACV. The outstanding electrochemical response may be due to the large surface area, excellent electrical conductivity, and remarkable electrocatalytic activity of rGO–TiO₂–Au nanocomposites, which provided the fastest electron transport at the electrode surface. These results clearly indicate that rGO–TiO₂–Au/GCE is very suitable for the determination of ACV.

Under optimized conditions (see Electronic Supplementary Material), LSV was recorded by varying the concentration of ACV at the rGO–TiO₂–Au/GCE (Figure 5). Figure 5A shows the LSV responses of rGO–TiO₂–Au/GCE for varying the concentration of ACV from 0 to 500 μM in 0.1 M PB solution. As illustrated, the successive increase of the peak current is observed by raising the concentration of ACV. The plots of peak current responses against ACV concentrations are shown in Figure 5B. It can be seen that the peak current responses of ACV increased linearly with its concentrations from 1 to 100 μM. For higher concentrations, a deviation from linearity is observed, which is due to the adsorption of ACV or its oxidation product on the electrode surface. The linear regression equation is Ipa (µA) = 0.1403c + 2.1814 (µM) with a correlation coefficient of 0.996 (Figure 5B inset). The detection limit is 0.3 μM based on a signal-to-noise ratio of 3. The rGO–TiO₂–Au/GCE showed a comparable or better dynamic range and detection limit for ACV compared to other electrochemical sensors, such as Cu nanoparticle-modified carbon paste electrode (2.64 μM) (Heli et al., 2010), pencil graphite electrode (0.3 μM) (Dilgin and Karakaya, 2016), nanoporous nickel microsphere-modified carbon paste electrode (40 μM) (Heli et al., 2012), fluorine-doped SnO₂ electrode (1.25 μM) (Martínez-Rojas et al., 2017), and CdO/Fe₃O₄-modified carbon paste electrode (0.3 μM) (Naghian et al., 2020).

To evaluate the reproducibility of rGO–TiO₂–Au/GCE for the analysis of ACV, five rGO–TiO₂–Au/GCEs were prepared independently with the same fabrication procedure, and their peak current values toward 100 µM ACV were compared. A relative standard deviation (RSD) of 3.18% was obtained, demonstrating the excellent reproducibility of the proposed sensor. The repeatability of rGO–TiO₂–Au/GCE was estimated by performing five successive measurements with the same rGO–TiO₂–Au/GCE in the same solution. It was observed that the oxidation peak current of ACV decreased continuously, which was due to the adsorption of an oxidative product of ACV on the electrode surface. In this case, rGO–TiO₂–Au/GCE can only be used for a single measurement. Meanwhile, the stability of TiO₂–Au–rGO/GCE was studied by storing the electrode in a refrigerator at 4°C. After two weeks, the peak current signal of the electrode remained at 90.2% of its initial response, suggesting the good storage stability of the modified electrode. The selectivity of rGO–TiO₂–Au/GCE toward the analysis of ACV was evaluated by testing some possible interfering species, including Na⁺, K⁺, Ca²⁺, Mg²⁺, Zn²⁺, Mn²⁺, glucose, sucrose, starch, dopamine, ascorbic acid, uric acid, guanine, paracetamol, and acetaminophen. As shown in Figure 6, a 100-fold concentration of each species had almost no influence on the peak current of ACV with deviations below 5%. These data revealed the good selectivity of rGO–TiO₂–Au/GCE toward the determination of ACV.

The practical application of rGO–TiO₂–Au/GCE in the analysis of real samples was studied by determination of ACV in commercial ACV tablet samples. The procedures for the analysis of tablets were followed as specified in Analysis of the Sample. The sample was spiked with three levels of ACV in the calibration range, and the content of ACV in the tablet samples was calculated by the standard addition method keeping the dilution factor in consideration. The result shows that the content of ACV was 196.2 mg per tablet, which was very close to the claimed value of 200 mg per tablet. Moreover, the recovery tests of ACV were performed, and the results are presented in Table 1. The satisfactory sample recoveries indicated the validity of the developed sensor for the determination of ACV in pharmaceutical formulations.