β-Cyclodextrin capped gold nanoparticles and nanodiamond assembled on a porous gold mini-chip as a disposable electrochemical sensor for bisphenol A

ABSTRACT Herein, a disposable electrochemical sensor was constructed from a gold DVD (GDVD) mini-chip modified with pores, β-cyclodextrin capped gold nanoparticles (AuNPs@β-CD), and nanodiamond (ND). For this purpose, the LSV method was employed to deposit a silver film on the surface of the mini-chip. Then, the silver was replaced with gold using galvanic replacement, after which etching of silver by nitric acid and ammonia was accomplished to obtain a porous GDVD mini-chip (PGDVD). Next, AuNPs@β-CD and thereafter, an aqueous ND dispersion were successively drop-cast on the PGDVD. Finally, the as-prepared mini-chip was exposed to air to obtain the PGDVD@AuNps@β-CD@ND mini-chip. The electrochemical oxidation behaviour of BPA showed an irreversible oxidative wave with decreased oxidation overpotential and increased current response for BPA. The calibration curve was linear over the concentration range of 5 × 10−5 M to 1 × 10−3 M of BPA. According to the 3sb/m and 10sb/m criteria, the limits of detection (LOD) and quantitation (LOQ) of the method were obtained to be 1 × 10−5 M and 5 × 10−5 M, respectively. Also, the reproducibility of the method (RSD%) was obtained to be 4%. The proposed sensor was employed for the analysis of spiked tap water and apple juice samples.


Introduction
Today, bisphenol A (BPA) or 2,2-bis(4-hydroxyphenyl) propane (Scheme S1) is considered as an industrial chemical building block in the manufacturing of polycarbonate plastics, such as disposable and reusable food storage containers, feeding bottles, sport safety equipment, and household electronics.In addition, it is used as a precursor to synthesise epoxy resin products including protective coatings used in metal cans [1,2].However, BPA is known as a hazardous endocrine disruptor compound that interfere with the endocrine function by mimicking hormones (estrogen) and has adverse health effects, leading to considerable impacts on both human body function and aquatic organisms [3][4][5].Moreover, BPA could be released from the synthetic materials into the environment by heat treatment or exposure to harsh detergents [6,7].Therefore, it is essential to establish a reliable technique to detect and control the level of BPA contamination.To date, various analytical methods have been developed for determination of BPA in real samples including gas chromatography coupled with mass spectrometry [8], liquid chromatography-tandem mass spectrometry [9,10], capillary electrophoresis [11], and fluorimetry [12,13].However, these techniques are costly, time-consuming, and demanding for elaborated sample pretreatment prior to the analysis step.On the other hand, electrochemical sensors possess the advantages of low cost, fast response, ease of miniaturisation, disposability, and simple operation [14,15].
BPA is an electroactive compound that reveals an electrochemical oxidative wave with rather low sensitivity using common commercial electrodes, such as glassy carbon [16,17] and gold electrodes [18][19][20].Consequently, modification of the electrode surface is necessary to achieve a sufficient sensitivity [21,22].More importantly, formation of oxidative polymeric products that result from the polymerisation of phenolic compounds can block the active sites of the electrode, leading to the fouling of the electrode surface [23][24][25][26][27][28][29][30].In order to deal with this problem, disposable electrochemical sensors provide the benefit of utilising a fresh sensing surface in each measurement run without the necessity of regeneration of the electrode surface [31].Furthermore, in comparison with common commercial electrodes, disposable sensors are more portable and inexpensive.
Throughout the recent years, nanomaterials have extensively drawn the attention of researchers for the construction of modified electrodes [32,33].Of all the wide variety of nanomaterials, carbon nanomaterials have over taken and passed the others in the field of the electrode modification due to their attractive properties, such as high chemical resistance, electrical, and thermal conductivity [34].Nanodiamond (ND), one of the carbon allotropes, consists of a compact tetrahedral sp 3 crystalline carbon core surrounded by an amorphous carbon shell [35][36][37][38].One of the greatest advantages of employing ND in the composition of modified electrodes is its wider working potential window with lower background current [39,40].
Owing to the excellent characteristics and shape-dependent properties including high chemical stability and biocompatibility, metallic nanoparticles (Au, Ag, Pt, Pd, Cu, etc.) have also found extensive applications in the various fields of biological and chemical sensing, catalysis, and optics [41,42].Among them, gold nanoparticles (AuNPs) have another outstanding feature of plasmonic properties.In addition, the large surface area of AuNPs favours its electrocatalytic activity [43][44][45].Several well-established methods have been reported for the preparation of AuNPs such as Brust and Turkevich methods [46,47].
In recent years, researchers are exploring the synthetic strategies that are environmentally conducive and eliminate the harmful substances.In this regard, the use of unmodified β-cyclodextrin (β-CD), as both eco-friendly reducing and capping agents, has been suggested [48,49].Cyclodextrins (CDs) are a class of supramolecules which consist of glucose units joined by α-1,4-glucosidic linkage (three most common types that are called α-, β-, and γ-CD made up of six, seven, and eight glucose units, respectively) and are prominent in the formation of host-guest inclusion complexes through non-covalent bindings.Due to the presence of primary and secondary hydroxyl groups in the narrow and wide sides of CDs, the outer surface is hydrophilic, while the glucose residues make the inner cavity hydrophobic.According to the size of the cavity, a variety of guest molecules or apolar moieties of the guests with different orientations could be accommodated into the cavity (from the wide or narrow side).Moreover, it is also feasible to form a complex by binding a molecule to the exterior part of CD [50,51].Furthermore, primary hydroxyl groups in the β-CD molecule can act as a reducing agent to reduce many metallic ions to their corresponding elemental forms.For example, β-CD can serve to reduce Au 3+ to Au, resulting in the formation of macrocyclic-AuNP organic-inorganic nanohybrids [52].In addition, it is noteworthy to mention that by providing an alkaline medium, the reducing power of β-CD is increased [53].
Nanoporous structured materials possess three-dimensional (3D) nano-sized pores and a broad range of applications that rely on their high permeability and large specific surface area [54].Particularly, nanoporous gold (NPG) is a good conductor and highly desirable for use in chemical sensing applications [55].There are several methods available for the preparation of NPG, involving anodising, templating, electrochemical deposition, galvanic replacement [56,57], and dealloying [58,59], or a combination of them [60,61].Electrochemical deposition is a facile, cost-effective, and adaptable route for the purpose of making solid films.Herein, the substrate plays in the role of the cathode.Another approach for the preparation of NPG structures is the galvanic replacement process.Thanks to its attractive characteristics involving versatility, simplicity, and spontaneous reaction, it is widely used for the production of noble metal nanoporous structures such as Au, Pd, Pt, etc.This method consists of two half reactions in which the less noble metal is corroded sacrificially, while the more noble metal (with higher reduction potential) ions in the solution is deposited at the surface.The driving force behind this reaction is the existence of a suitable reduction potential difference between the two metal species [62,63].On the other hand, Erlebacher et al. investigated the evolution of porosity in the Ag-Au alloy process, and proposed the dealloying process as a generic method for preparation of NPG [64].During the chemical dealloying, the more active (less noble) metal is dissolved by immersing the binary alloy in a selective etchant solution [65].
The aim of this work is to introduce a novel disposable electrochemical sensor based on a PGDVD mini-chip modified with AuNPs@β-CD and ND for the detection and determination of BPA in tap water and apple juice samples.Herein, the PGDVD mini-chip, as the substrate, provides an enhanced surface area and thus, higher electrocatalytic activity.Also, β-CD plays a central role as a benign reducing agent and also as a capping agent for the preparation of AuNPs@β-CD, as well as a host for the accommodation of the hydrophobic BPA guest.Plus, ND could diminish the background current and noise to improve the limit of detection of the method.A stock solution of BPA (0.1 M) was prepared in ethanol and preserved at 4°C.In addition, Britton-Robinson (BR, 0.04 M; H 3 BO 3 /CH 3 COOH/H 3 PO 4 ,) and phosphate (PBS, 0.1 M; KH 2 PO 4 /K 2 HPO 4 ) buffer solutions with various pH values were used as electrolyte solutions.
The morphological and topological images of the working electrode surface were obtained using a scanning electron microscope (SEM, LEO 1450 VP, Germany) and an atomic force microscope (AFM, Ara Pajoohesh, full plus, Iran), respectively.Moreover, the elemental analysis of samples were obtained by means of Energy-Dispersive X-ray spectroscopy (Oxford, 7353, England).Also, in order to estimate the size of the nanomaterials, transmission electron microscopy (TEM, LEO, 912AB, Germany) and dynamic light scattering (DLS, Cordouan, Vasco, France) were used.Absorption spectrum in the UV-Vis region was recorded using a single beam UV-Vis spectrophotometer (Agilent, 8453, USA).A zeta potential analyser (CAD, Zeta Compact, France) was used to determine the zeta potential of synthesised gold nanoparticles.A potentiostat/galvanostat (EIS, WonAtech, ZIVE SP1, Korea) with ac voltage of 10 mV and frequency range of 0.1 to 100 kHz was used to measure electrochemical impedances.

Electrochemical measurements
10 mL of the supporting electrolyte solution was transferred to a clean and dry glass cell and the required amount of the stock BPA solution was added to the solution using a micropipette.After stirring well, the electrochemical experiments were performed using LSV, CV or CHP electrochemical technique.

Fabrication of GDVD mini-chip
DVD-R (Archival Gold® Delkin Devices Inc., Poway, CA, USA) was purchased from the local stores.At first blush, the DVDs were cut into pieces of 10 mm in 30 mm using a laser cutter machine.Next, in the laboratory, a mechanical force was exerted by hand with the aid of a cutter, which easily removed the protective polycarbonate layer such that the thin-gold layer of the mini-chip was exposed.Then, the gold surface was cleaned by immersing the mini-chip in pure ethanol for 10 minutes and washing thoroughly with deionised water.In order to provide the electrical contact, a strip of copper was attached to the end of the mini-chip.Then, a cold laminate adhesive with a 3 mm diameter hole was used to cover the gold surface completely.In order to assure the electrical insulation, the electrode was wrapped up in electrical tape.All the required steps for the construction of the GDVD mini-chip are shown in Scheme S2.

Synthesis of AuNPs@β-CD
The AuNPs@β-CD were prepared simply as follows: First, 0.014 g of β-CD solid powder was added to 100 μL of 0.01 M tetrachlorauric acid solution.Next, 4.9 mL of deionised water was added to the mixture and then, the pH of the solution was adjusted to 10.5 by the gradual addition of 1 M NaOH solution.Finally, the solution was placed in a water bath (60°C) for 3 h, while being stirred vigorously.During the synthesis procedure, the solution colour changes from colourless to pink [48,53].The resulting AuNPs@β-CD dispersion was stable under the storage conditions (at ~4°C) and remained unchanged for several months.

Preparation of ND dispersion
In order to prepare an aqueous nanodiamond dispersion, 1 mg of ND was added to 1 mL of deionised water [36].Henceforth, the as-prepared mixture was ultrasonicated for 1 h in room temperature.

Preparation of the PGDVD@AuNPs@β-CD@ND modified mini-chip
A series of sequential modification processes was performed to prepare the PGDVD@AuNPs@β-CD@ND modified mini-chip, which is as follows:

Fabrication of PGDVD mini-chip
mini-chip, several steps have to be performed.Firstly, a thin Ag film was electrodeposited at the surface of the GDVD mini-chip from an electrolyte solution containing 0.01 M Ag 2 SO 4 , 1.5 M KSCN, and 0.75 M NH 4 Cl using LSV technique.The applied potential range was from +0.2 to −1 V (vs.saturated Ag/AgCl electrode) and the scan rate was 1 mV/s.The Ag film could be alternatively electrodeposited by the means of CHP technique using the same electrolyte solution with the applied current of 8 mA for 30s.The as-prepared modified mini-chip-is symbolised as GDVD@Ag.Secondly, galvanic replacement process was accomplished to replace the most of the Ag atoms with the Au atoms.For this purpose, the GDVD@Ag minichip was soaked in a 0.05 mM HAuCl 4 solution in the ambient conditions for 12 h [66] and then, washed thoroughly with the deionised water.The as-prepared modified mini-chip is called GDVD@AgAu.Thirdly, a dealloying process was carried out to remove the remaining unreacted Ag atoms and to obtain a porous gold film.This was done by the selective dissolution of Ag atoms from GDVD@AgAu minichip in a 7.5 M nitric acid solution, followed by rinsing the mini-chip with the deionised water.Finally, the mini-chip was immersed in a 25% w/w ammonia solution to completely remove the residual Ag and AgCl precipitates [66] and again, it was thoroughly washed with the deionised water.Such a mini-chip is named PGDVD.

Construction of PGDVD@AuNPs@β-CD modified mini-chip
Initially, 5 μL of the AuNPs@β-CD dispersion (as prepared in section 2.5) was drop-cast on the surface of the PGDVD mini-chip using a 10 μL microsyringe.In the next stage, the mini-chip was left for 1 h at room temperature to permit the solvent evaporation.The provided mini-chip is called PGDVD@AuNPs@β-CD.

Preparation of PGDVD@AuNPs@β-CD@ND modified mini-chip
In this step, 10 μL of the ND suspension (as prepared in section 2.6) was carefully drop-cast on the surface of PGDVD@AuNPs@β-CD mini-chip that was subsequently left to be dried under ambient conditions for 2 days.This mini-chip is characterised as PGDVD@AuNPs@β-CD@ND.The whole procedure for the preparation of the modified mini-chip is shown in Scheme 1.

Real sample preparation
A tap water sample (from the laboratory) and an apple juice sample without pulps (from the local markets) were first diluted appropriately with phosphate buffer (0.1 M, pH 7.0).The tap water sample was then filtered using a Whatman® filter paper (No. 40) to remove insoluble particles.Scheme 1. Schematic representation of the procedure for the preparation of the modified PGDVD@AuNPs@β-CD@ND mini-chip.

Characterisation of ND
The TEM image of the aqueous dispersion of ND, shown in Fig. S1, illustrates that the particles are cluster-like with the individual particle size of ~10 nm.

Synthesis and characterisation of AuNPs@β-CD
Unmodified β-CD molecules show a bifunctional performance in the synthesis of AuNPs@β-CD nanoparticles.Under mild conditions, β-CD acts as both reducing and capping agents [48,67].On the one hand, the hydroxyl groups of β-CD act as reducing agents, which reduce the Au 3+ into the metallic Au, while being oxidised to carboxylic acid moieties themselves.On the other hand, they also act as capping agents that surround and stabilise the freshly prepared AuNPs, leading to the formation of AuNPs@β-CD.Interaction of the carboxylate anions with AuNPs results in the formation of Au-COO − groups at the surface of AuNPs [68,69].
Herein, the formation of AuNPs@β-CD was confirmed by UV-vis absorption spectroscopy.Figure 1a shows the UV-vis absorption spectrum of the synthesised AuNPs@β-CD in the range of 300 to 800 nm.It is observed that the colloidal pink solution of AuNPs@β-CD shows the characteristic absorption band at λ max = 524 nm, which corresponds to surface plasmon resonance (SPR) of AuNPs and it differs from the absorption band of β-CD (λ max = 220 nm, not shown here) [48,70].Moreover, the synthesised nanoparticles were characterised by dynamic light scattering (DLS), which is also known as photon correlation spectroscopy (PCS).The technique is used to analyse the hydrodynamic size of the nanoparticle suspensions.Figure 1b displays the results of DLS analysis.It is realised that the range of diameters of the AuNPs is from 10 to 40 nm and the average particle size is about 19 nm.Furthermore, to investigate the morphology, distribution, and uniformity of AuNPs@β-CD nanoparticles, the TEM image of the AuNPs@β-CD colloidal solution was obtained and indicated in Figure 1c.It is obvious that the synthesised nanoparticles are uniform and spherical in shape.Also, their sizes are in the range of 12 to 19 nm, as had been shown by DLS.In addition, the zeta potential of AuNPs@β-CD was measured to investigate the stability of the colloidal systems.It should be noted that the colloidal systems with high zeta potential, either positive or negative, are stabilised electrically.
According to the obtained zeta potential data (Figure 1d), the average zeta potential of AuNPs@β-CD is −33 mV that indicates the stability of AuNPs@β-CD colloidal solution.

Characterisation of the PGDVD@AuNPs@β-CD@ND mini-chip
SEM Studies.Fig. S2a shows the SEM image of a bare GDVD mini-chip surface, which exhibits a series of regular and parallel corrugations.As expected, the presence of Au on the surface of GDVD electrode is quite obvious from the EDS spectrum in Fig. S2b.
In order to construct the GDVD@Ag modified mini-chip, an appropriate potential or a potential range must be applied to the GDVD mini-chip to trigger the reduction of Ag + at the surface of the GDVD.Herein, CHP and LSV techniques have been employed for the electrodeposition of Ag at the GDVD.Figs.S3a,b displaythe SEM images of GDVD@Ag mini-chip surfaces prepared by CHP and LSV techniques, respectively.It is obvious that the CHP results in the accumulation of Ag in a limited area of the GDVD surface, whereas LSV method brings about a more uniform and homogeneous electrodeposited Ag film, showing that the LSV is a more efficient electrochemical technique for the electrodeposition of the Ag film.Furthermore, the EDS spectrum of the GDVD@Ag mini-chip (prepared by LSV method) is illustrated in Fig. S3c in which the presence of Ag at the surface of the modified electrode after the electrodeposition process is confirmed.Also, it should be noted that the presence of K is attributed to the presence of KSCN in the electrolyte solution.
In the next step, the spontaneous dynamic galvanic replacement process was accomplished to prepare the GDVD@AgAu modified mini-chip.Figs.S4a,b indicate the SEM image and the EDS spectrum of the GDVD@AgAu mini-chip, respectively.From the SEM images, a distinct change in the mini-chip surface structure is detectable after the galvanic replacement process.Also, the EDS spectrum reveals that after the galvanic replacement, the amount of the Ag has been diminished and on the contrary, the amount of the Au has been enhanced at the mini-chip surface.
AFM Studies.Since the porosity of the electrode surface plays a significant role in its electrocatalytic activity, surface topography should be considered during modifying the electrode.AFM as a type of high-resolution scanning probe was employed to show the topography of the mini-chip surface during its modification process (Fig. S5).Fig. S5a shows the unmodified surface of a bare GDVD mini-chip whose grooves are clearly apparent.Moreover, Fig. S5b demonstrates the AFM image of the GDVD@Ag modified mini-chip from which it is realised that upon Ag electrodeposition, the gold grooves are covered with Ag.The corresponding electrochemical process for this stage can be shown as Eqn. 1 [71,72] Furthermore, Fig. S5c belongs to the GDVD@AgAu modified mini-chip in which the formation of AgAu alloy during the galvanic replacement process has partially reduced the porosity of the electrode surface.As the standard reduction potential of the AuCl 4 − /Au pair (+0.99 V vs. SHE) is higher than that of the Ag + /Ag redox couple (+0.80 V vs. SHE), Ag atoms act as sacrificial templates, which are oxidised and dissolve to form AgCl; whereas, the AuCl 4 − ions are reduced and deposit as Au atoms at the electrode surface.The galvanic replacement process can be shown as Eqn. 2 [66]: In addition, Fig. S5d refers to the surface topography of the PGDVD mini-chip in which several pitches are observed after the dealloying process.In this stage, the residual Ag atoms and AgCl are removed by immersing the GDVD@AgAu modified mini-chip in the etchant solutions of HNO 3 and NH 4 OH, successively [73,74].As a result, the dealloying of AgAu occurs and a thin porous Au film remains on the surface.The corresponding reactions are shown in Eqn. 3 and Eqn. 4 [75,76].
Also, Fig. S5e shows the surface topography of the PGDVD@AuNPs@β-CD mini-chip that has been prepared by the drop-casting of the AuNPs@β-CD dispersion on the PGDVD mini-chip surface.It is seen that the AuNPs@β-CD nanoparticles have been evenly distributed and the porosity of the surface has been clearly enhanced.At last, the AFM image of the PGDVD@AuNPs@β-CD@ND mini-chip is shown in Fig. S5f, from which it can be found that the drop-cast ND helps to improve the active surface area and the porosity of the modified mini-chip.EIS Studies.EIS is a useful tool to investigate the effect of coating layers on the electron transfer kinetics [77].The Nyquist plot, in which the data appears to be in the form of real impedance in terms of imaginary impedance, is the most suitable method for presenting electrochemical apparent impedance data of a circuit at various frequencies.Fig. S6 represents the Nyquist plots for the GDVD, GDVD@AuNPs@β-CD, GDVD@ND, GDVD@AuNPs@β-CD@ND, and PGDVD@AuNPs@β-CD@ND mini-chips, which have been obtained in a solution of 1 mM [Fe(CN 6 )] 3-/4-and 0.1 M KCl at the open-circuit potential (OCP).Plus, the inset of Fig. S6 shows the Randles electrical equivalent circuit (EEC) with which the data have been fitted.Comparing the semi-circle diameter of curve a with that of curve b in Fig. S6 reveals that the R ct has decreased from 3095 Ω to 2349 Ω after the modification of the GDVD mini-chip with AuNPs@β-CD, indicating an increase in the electron transfer rate at the electrode-electrolyte interface, which could be attributed to the electrocatalytic activity of AuNPs [78,79].On the other side, when the GDVD is modified with ND, a considerable increase in R ct from 3095 Ω (in curve a) to 4699 Ω (in curve c) is observed, which can be related to the weak conductivity of ND.Moreover, a decrease in R ct is observed when curve d is compared with curve b, showing that the modification of GDVD@AuNPs@β-CD with ND cause the R ct to be unexpectedly reduced from 2349 Ω to 1980 Ω.This observation can be ascribed to the synergistic interaction between AuNPs@β-CD and ND, creating diverse electrocatalytic activities.Eventually, in the case of the PGDVD@AuNPs@β-CD@ND mini-chip (curve e), a straight line is observed which implies that the R ct has dramatically decreased from 1980 Ω to 600 Ω when compared to the GDVD@AuNPs@β-CD@ND.Apparently, this could be due to the extensively enlarged surface area of the PGDVD in comparison with the pristine GDVD.

Electrochemical Behaviour of BPA at diverse mini-chips
To elucidate the electrochemical behaviour of BPA at diverse modified electrodes, cyclic voltammetry experiments have been performed using each of the modified mini-chips in 0.1 M PBS (pH 7.0) containing 1 mM BPA within the potential range of 0.0 to +0.8 V with a scan rate of 50 V/s and the corresponding voltammograms have been shown in Figure 2. From curve a, it is realised that BPA on the GDVD mini-chip shows an oxidation peak in the anodic scan at the peak potential (E p ) of +0.54 V and no reduction peak in the reverse scan, which implies BPA oxidation is a slow irreversible charge transfer process.Moreover, curve b, corresponding to the PGDVD mini-chip, indicates that the peak current (I p ) for BPA oxidation has considerably increased compared to the GDVD, while no significant change in E p is observed.This could be attributed to the significant enhancement of the active surface area caused by porosity, which accelerate the electron transfer rate for BPA oxidation, resulting in approximately 5-fold increase in the I p .In addition, with the PGDVD@AuNPs@β-CD mini-chip (curve c), a further increase in the I p (34.54 μA) is observed that can be related to the excellent electrocatalytic activity of AuNPs.On the other hand, the E p , located at +0.647 V, has shifted towards more positive potentials, which is probably due to the formation of an inclusion complex between β-CD and BPA as a result of their host-guest interaction.Furthermore, curve d, corresponding to the PGDVD@AuNPs@β-CD@ND mini-chip, shows that the I p (43.91 μA) has dramatically increased compared to all of the previously studied ones.Also, the E p , occuring at +0.567 V, has considerably shifted towards less positive potentials in comparison with the PGDVD@AuNPs@β-CD mini-chip.These phenomena, the increment in the I p and decrement in the E p , suggest an excellent electrocatalytic activity of the mini-chip towards oxidation of BPA.Plus, curve e shows that with the PGDVD@AuNPs@β-CD@ND mini-chip no redox peak is observed when BPA is absent.

Optimisation of the electrochemical sensor performance
In order to fabricate an efficient electrochemical sensor, several factors affecting the sensor performance were thoroughly investigated and optimised.These factors are classified into two major groups including fabrication factors, such as the time intervals of the galvanic replacement and dealloying process as well as operational factors, such as the type of the supporting electrolyte and the pH of the solution.

Optimisation of the fabrication factors
Time Interval of the Galvanic Replacement.The influence of the required time interval for galvanic replacement process on the performance of the sensor was studied.Various time intervals, including 6, 12, and 18 h, were considered.The results showed that the time interval of 6 h is not adequate for the complete replacement of Ag, while 18 h would cause most of the mini-chips to be damaged due to the liquid leaking into their connections and swelling.On the other hand, 12 h galvanic replacement in the presence of 50 μM HAuCl 4 solution leads to the highest peak current and the least positive potential values.
Dealloying Time.Dealloying is necessary to remove the unreplaced Ag atoms and obtain a porous gold film.With the intention of dealloying, the GDVD@AgAu mini-chip is first placed in a 7.5 M nitric acid solution.The procedure is followed by immersing the mini-chip in an ammonia solution (25% w/w) for 5 min to completely remove the residual Ag and AgCl precipitates.The immersion time of the GDVD@AgAu mini-chip in 7.5 M nitric acid solution dramatically affects the efficiency of dealloying process and the sensitivity of the sensor.Fig. S7, demonstrates that the Ag oxidation peak is completely eliminated within 10 minutes of immersion, indicating the complete removal of the unreplaced Ag atoms.

Optimisation of the operational factors
Effect of pH.The effect of pH on the electrochemical behaviour of BPA at the PGDVD@AuNPs@β-CD@ND mini-chip was studied using cyclic voltammetry.A series of BR buffers with different pHs in the range of 3 to 9 were used as supporting electrolyte to ensure the same matrices in all the measurements.The obtained voltammograms, shown in Fig. S8, reveals that with increasing pH, the anodic peak potential (E p ) shifts towards less positive potentials, indicating a facilitated electron transfer.On the other hand, the anodic peak current (I p ) reaches its maximum at pH 7.0.Therefore, the optimum value for pHwhich has been utilised for future experiments was 7.
Selection of the Supporting Electrolyte.It is well-known that the supporting electrolyte is an important effective factor for electrochemical experiments and selection of a proper medium can influence the electrochemical response.For the purpose of finding an appropriate electrolyte, electrochemical responses of BPA were investigated in BR (0.04 M) and phosphate buffers (0.1 M), both in pH 7.0, at the PGDVD@AuNPs@β-CD @ND mini-chip.As depicted in Fig. S9, comparison of the two buffers shows that the background current is much higher using BR buffer.Instead, from the CV plots in Fig. S10, it is revealed that the I p for BPA is slightly higher in PBS and the E p occurs at less positive values.Therefore, PBS was selected as the optimum supporting electrolyte.
Effect of Scan Rate.The effect of scan rate on the voltammetric response of BPA on the PGDVD@AuNPs@β-CD@ND mini-chip was studied over the range of 10 to 450 mV/s in order to clarify the reaction mechanism.Fig. S11a shows the clear dependence of the oxidation E p on the scan rate along with the absence of the reduction peak in the reverse scan, which confirms the irreversibility of BPA oxidation on the surface of the mini-chip.Moreover, as demonstrated in Fig. s11b, the plot of I p versus the scan rate is linear, indicating an adsorption-controlled process for the reaction.The linear regression equation as declared in Eqn. 5 is: In which, ν refers to the scan rate.Also, as shown in Fig. S11c, a linear correlation can be observed between the E p and the logarithm of the scan rate for which the regression equation has obtained to be as Eqn.6: According to Laviron's model, for a totally irreversible and adsorption-controlled process, the slope of the Eqn.6 can be calculated using Eqn.7 [80]: n denotes the total number of the involved electrons, F is the Faradaic constant, and α is the transfer coefficient.For an irreversible reaction, α is supposed to be 0.5; so that the total number of the involved electrons has been calculated to be 1.49, which is rounded off to 1. Now that the number of electrons is determined, the exact value of α can be calculated from Eqn. 7 by placing the number of electrons equal to 1.In this manner, the value of α is obtained to be 0.25.A possible one electron oxidation route is presented in Scheme S3 [81,82].

Fouling of the electrode surface
In order to investigate the fouling of the electrode surface upon the oxidation of BPA, four consecutive scans were performed.As shown in Fig. S12, a remarkable decline in the BPA anodic peak current is noticed, such that it almost disappears in the fourh scan.This phenomenon could be attributed to the fouling or passivation of the mini-chip surface by oxidative polymeric products that blocks the sensing layer and prevents the BPA molecules from reaching the surface.Due to such common fouling behaviour of BPA oxidation, designing and fabrication of a disposable electrochemical sensor for BPA seems necessary.

Interference study
The impact of several possible organic (glucose, citric acid, ascorbic acid, 1-naphtol, and 4-aminophenol) and inorganic (Na + , Mg 2+ , Ca 2+ , SO 4 2-, NO 3 − , and Cl − ) concomitants on the voltammetric response of the proposed mini-chip towards BPA oxidation were examined (Fig. S13).Table 1 indicates the influence of concomitant species on BPA anodic I p as the recovery percentage.Considering ±5% as the tolerable and acceptable error, Na + , NO 3 − , glucose, and citric acid showed no interference towards determination of BPA at 1:1 molar ratio, while other species were interfering.

Analytical performance
In order to investigate the analytical performance of the proposed BPA sensor, the analytical figures of merit were obtained.Fig. S14 shows the corresponding calibration plot (I p versus BPA concentration) which was obtained using LSV method under the optimised operating conditions.The linear range was from 5 × 10 −5 M to 1 × 10 −3 M of BPA with a regression equation of I p (μA) = 0.039502 C BPA (μM) + 6.3669 and a squared correlation coefficient (R 2 ) of 0.9939.The limit of detection (LOD) and the limit of quantification (LOQ), based on 3s b /m and 10s b /m criteria, were found to be 1 × 10 −5 M and 5 × 10 −5 M, respectively.Moreover, the reproducibility (RSD%) of the method, evaluated by six successive measurements with similar electrodes, was obtained to be 4%.

Real sample analysis
In order to evaluate the applicability of the proposed sensor for determination of BPA in real samples, the PGDVD@AuNPs@β-CD@ND mini-chip has been employed for the measurement of BPA amount in tap water as well as glass-bottled apple juice samples.After the preparation of the samples according to the procedure described in section 2.8, they were spiked by known amount of BPA and analysed using the standard addition method.The results shown in Table 2 indicate good recoveries according to ±5% acceptable error.

Conclusion
In summary, a novel modified GDVD mini-chip was successfully fabricated as an electrochemical sensor for the determination of BPA in real samples.The special interest in this work was the use of GDVDs as disposable mini-chips in order to tackle the problem of electrode surface fouling leading to the poor repeatability.A facile multi-step method, which was based on electrodeposition, galvanic replacement, and alloying/dealloying processes, enabled the authors to produce efficient porous GDVD mini-chips with high surface-to-volume ratio.Next, further modification was made by the drop-casting of ~19 nm average-sized AuNPs@β-CD that were directly synthesised using environmentally-friendly unmodified β-CD, playing both reducing and stabilising agents and providing host-guest interaction with BPA molecule.Finally, a ND dispersion was drop-cast to prepare the PGDVD@AuNPs@β-CD@ND modified mini-chip.The large specific surface area created by the incorporation of the porous structure of the PGDVD, the high electrocatalytic activities and the synergistic effects of AuNPs@β-CD and ND, and also supramolecular recognition properties of β-CD for BPA as a guest molecule led to significant signal amplification.The proposed disposable electrochemical senor shows merits of cost-effectiveness, improved sensitivity, and good reproducibility.

Table 1 .
Results of the interference study (BPA concentration of 1 mM).

Table 2 .
Results of quantitative determination of BPA in real samples using the proposed electrochemical sensor.