Introduction

Folic acid (FA), a B vitamin, is water soluble and has a major role in biological functions [1]. Folate deficiency causes megaloblastic anemia and subclinical deficiency, chronic liver disease, biochemical and hematologic lesions, and birth defects [2]. In contrast, high doses of FA during pregnancy are associated with low weight for gestational age at birth [3]. Therefore, the rapid and accurate measurement of FA and the assessment of its folate nutritional status are really important. Liquid chromatography–ion chromatography (LC-IC) [4], fluorescent probe [5], LC–tandem mass spectrometry (LC-MS/MS) [6], surface plasmon resonance [7], and so on have been developed for the detection or determination of FA. In particular, electrochemical methods have attracted increasing attention due to their advantages such as speed, precision, lower reagent usage, and lower equipment costs compared with certain analytical techniques [8, 9]. For example, cobalt oxide–reduced graphene oxide–cetyltrimethyl ammonium bromide modified carbon paste electrode (Co3O4-rGO-CTAB/CPE) was fabricated for sensing FA in real samples by differential pulse voltammetry (DPV) [10]; three-dimensional (3D) rGO/FeNi3–ionic liquid nanocomposite modified CPE showed excellent electrocatalytic properties towards FA owing to the synergistic effects of the two components [11]; carbon black (CB), GO, copper nanoparticles (CuNPs), and poly(3,4-ethylenedioxythiophene)-poly(styrenesulfonate) (PEDOT:PSS) composite was prepared by ultrasonic agitation and used as electrode modifier for detection of FA by square-wave voltammetry (SWV) [12]; graphene/ZnO nanowire arrays/graphene foam modified glassy carbon electrode was used for determination of FA in tablets and urine [13]. Thus, the development of fast and facile sensing systems for the measurement of FA in real samples is an ongoing, active area of research [14].

Cyclodextrins (CDs) are a well-known series of cyclic oligosaccharides containing 6–12 d-glucopyranose units [15]. CDs exhibit some special properties such as a hydrophobic cavity of different sizes, natural chirality, and external hydrophilicity. Many organic and inorganic compounds can be embedded into the hydrophobic cavity of CDs [16]. To some extent, the physicochemical properties of the encapsulated materials can be changed a lot [17]. That is to say, functional CD-based molecular recognition can be obtained owing to the possibility of formation of inclusion complexes between inorganic/organic substances and CDs; this has also been beneficial to their wide application in electrochemical recognition [18, 19]. β-CD containing seven glucopyranose units with a cavity size suitable for accommodating aromatic rings and it readily forms an inclusion complex (IC) with FA [20]. β-CD is widely used as a dispersing agent for insoluble chemicals and nanomaterials [21]. In addition, nanocomposites composed of β-CD also exhibited high stability and could be used in highly efficient electrochemical recognition due to the synergistic effects of the nanoparticles and β-CD [22].

Being composed of carbon atoms which form one-dimensional (1D) systems with nanostructures, carbon nanotubes (CNTs) have shown some specific properties such as high thermal stability, excellent conductivity, huge surface area, and relatively high mechanical and chemical capability. CNTs are classified into two types, multi-walled CNTs (MWCNTs) and single-walled CNTs (SWCNTs), according to the number of carbon atom layers [23]. CNT-based sensors were developed and they have many advantages such as high sensitivity and selectivity. Fe3O4/multi-walled carbon nanotubes (Fe3O4/MWCNTs) and Fe3O4/fluorinated MWCNTs (Fe3O4/F-MWCNTs) nanocomposite modified electrodes were used for the sensitive, selective, and simultaneous electrochemical detection of multiple heavy metals in the environment [24]; a nanostructured silver-doped TiO2/CNTs hybrid-based sensor was used as a sensitive sensor for electrochemical detection of cetirizine [25]; an MWCNT-SiO2@Au core-shell nanocomposite-based sensor was used for ultrasensitive detection of bisphenol A, indicating that MWCNTs could be used as efficient electrode modifiers [26].

Graphene, a sheet-like structure containing one or multiple layers of six-membered carbon rings [27, 28], possesses a two-dimensional (2D) honeycomb lattice structure. Up to now, graphene has become a basic unit of many novel nanocomposites. The carbon atoms on the graphene are sp2 hybridized, which contributes a lot to its relatively high conductivity [29]. Various graphene oxide (GO) or graphene-based sensors were developed for selective electrocatalytic recognition [30,31,32]. Novel composite materials fabricated by the combination of graphene or GO or graphene nanoplatelets (GNPs) with noble metal nanoparticles (NMNPs) such as gold nanoparticle (AuNPs) [33], platinum nanochains (PNCs) [34], platinum nanoparticles (PtNPs), and so on have been shown to be very effective.

Herein, a PtNPs-GNPs-MWCNTs-β-CD nanocomposite was fabricated through a facile ultrasound-assisted assembly method. The nanocomposite combines the advantages of β-CD, PtNPs, GNPs, and MWCNTs such as good conductivity, large specific surface area, and excellent inclusion properties. The electrocatalytic properties of PtNPs-GNPs-MWCNTs-β-CD nanocomposites modified carbon glass electrode (PtNPs-GNPs-MWCNTs-β-CD/GCE) toward FA was evaluated by cyclic voltammetry (CV) (Scheme 1). The application of PtNPs-GNPs-MWCNTs-β-CD/GCE in detection of FA in real samples was also investigated.

Scheme 1
scheme 1

Fabrication of PtNPs-GNPs-MWCNTs-β-CD/GCE and its electrochemical behavior toward FA

Full size image

Materials and methods

Materials and apparatus

FA (99 wt%) and β-CD (B.R.) were purchased from Tokyo Chemical Industry Co, Ltd. (Shanghai, China). Chloroplatinic acid hexahydrate (H2PtCl6·6H2O, 99.95 wt%), sodium citrate, and glutamate were provided by Aladdin Industrial Co., Ltd. (Shanghai, China). GNPs (99.5 wt%) and MWCNTs (> 95 wt%) were provided by Nanjing Xianfeng Nano Technology Co., Ltd. (Nanjing, China). Potassium ferricyanide (K3[Fe(CN)6], ≥ 99.5% wt%) was obtained from Tianjin Broadcom Chemical Co., Ltd. (Tianjin, China). Disodium hydrogen phosphate dodecahydrate (Na2HPO4·12H2O, ≥ 99.5 wt%), sodium dihydrogen phosphate dihydrate (NaH2PO4·2H2O, ≥ 99.5 wt%), phosphoric acid (H3PO4, 98 wt%), sodium hydroxide (NaOH ≥ 96.0 wt%), sodium chloride (NaCl, ≥ 99.5 wt%), potassium chloride (KCl ≥ 99.5 wt%), magnesium chloride (MgCl2, ≥ 98.0 wt%), zinc nitrate (Zn(NO3)2, ≥ 99.0 wt%) and calcium chloride (CaCl2, ≥ 96.0 wt%) were purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Glucose (C6H12O6 H2O, A.R.) was purchased from Tianjin Heng Xing Chemical Co., Ltd. (Tianjin, China). Folic acid tablets were obtained from Tianjin Feiying Pharmaceutial Co., Ltd. (Tianjin, China). Alumina polishing powder (Al2O3, 0.05 μm) was purchased from Lab Testing Technology Co., Ltd. (Shanghai, China). All of the solutions were freshly prepared using high-purity water with an electrical resistivity of 18.2 MΩ cm, which was produced by a Millipore Milli-Q water purification system (Millipore Trading Co., Ltd.; Shanghai, China). A thermal heating magnetic mixer (DF-101S; Gongyi Yuhua Instrument Co., Ltd., China), a centrifugal machine (TGL-16GA; Zhengzhou Changcheng Science Industry & Trade Co., Ltd., China), and an ultrasonic cleaner with an output power of 250 W (SB-100D; 40 kHz; Ningbo Scientz Biotechnology Co., Ltd., China) was used to fabricate PtNPs and the PtNPs-GNPs-MWCNTs-β-CD composite.

Synthesis of PtNPs

PtNPs with average particle sizes of 3.5–5.2 nm were prepared according to the previously reported method [35]. Thus, 100.0 mL of H2PtCl6·6H2O (1.0 mmol L−1) solution was heated to 110 °C, and then 10.0 mL of aqueous sodium citrate (38.8 mmol L−1) was added into the reaction. After 30 min, the solution turned black. The solution was stirred for another 10 min and then allowed to cool to room temperature. The mixture was rinsed twice with high-purity water and centrifuged, and the residual solid was dispersed in 25.0 mL of high-purity water to obtain aqueous PtNPs dispersion, which was stored in a refrigerator at 4 °C for future use.

Fabrication of PtNPs-GNPs-MWCNTs-β-CD composite modified electrode

A homogeneous dispersion of 4.0 mg of GNPs, 4.0 mg of MWCNTs, and 2.0 mL of PtNPs was prepared by 250-W ultrasound treatment in ultrapure water. After 100 μL of aqueous β-CD (2.0 mg mL−1) solution was added into the reaction, it was sonicated for another 10 min to obtain assembled PtNPs-GNPs-MWCNT-β-CD composite.

Bare GCE was polished with alumina (0.05 μm) slurries and then sonicated in water for 3 min. After it was dried in a stream of nitrogen (N2), PtNPs-GNPs-MWCNTs-β-CD suspension with different volumes was dropped onto the surface of the GCE and allowed to dry to obtain PtNPs-GNPs-MWCNTs-β-CD/GCE.

Electrochemical measurements

Each sample was measured three times consecutively to evaluate the reproducibility of the modified electrodes, and relative standard deviation (RSD = standard deviation/mean × 100%, n = 3) was calculated. Electrochemical experiments were carried out on a CHI660E Electrochemical Workstation (Chenhua Instrument Co., Ltd.; Shanghai, China) with a classical three-electrode system which contained a Pt auxiliary electrode, an Ag | AgCl | KCl(sat.) reference electrode, and a working electrode.

Characterization

The microstructure and morphology of the samples were characterized using a high-resolution transmission electron microscope (HR-TEM, JEM-2100F, JEOL Ltd., Tokyo, Japan), a field emission scanning electron microscope (FE-SEM, TESCAN MIRA3 LMH/LMU, Czech), and an atomic force microscope (AFM, Bruker Dimension Icon, operating in tapping mode, USA). The functional groups of materials were also characterized by Fourier transform infrared spectroscopy (FT-IR, Perkin Elmer Spectrum 100, USA). Raman spectra were recorded at room temperature with a confocal micro-Raman spectrometer (Horiba LabRAM HR 800, UK).

Results and discussion

Characterization of materials

Morphological analyses

The morphology of PtNPs was investigated by HR-TEM. As shown in Fig. 1, the PtNPs were monodisperse with particle sizes of 3.5–5.2 nm. Furthermore, ordered lattices of PtNPs could also be found. As shown in Fig. 1b, the d spacing of adjacent fringes for PtNPs is 0.227 nm, which corresponds to the (111) crystal plane of face-centered cubic (fcc) Pt phase (Fig. 1b) [36, 37].

Fig. 1
figure 1

HR-TEM images of PtNPs. a Low resolution. b High resolution

Full size image

The morphology of PtNPs-GNPs-MWCNTs-β-CD was observed by FE-SEM. Figure 2 shows that the PtNPs-GNPs-MWCNTs-β-CD composite possesses a loose, porous, and rough structure due to the ultrasound-assisted assembly of PtNPs, GNPs, MWCNTs, and β-CD. It is proposed that more effective electron transfer between a modified electrode with porous interface and an analyte could be facilitated, thereby greatly improving the electrochemical responses [38].

Fig. 2
figure 2

FE-SEM images of PtNPs-GNPs-MWCNTs-β-CD composite. a Low resolution. b High resolution

Full size image

The surface characteristics of the developed sensor were investigated by using AFM. As shown in Fig. 3a, bare GCE possessed a homogeneous structure and low-roughness features (3.87 nm). Compared to bare GCE, a 100-fold dilution of the PtNPs-GNPs-MWCNTs-β-CD composite modified GCE showed higher surface roughness (7.05 nm) (Fig. 3b), again indicating that the PtNPs-GNPs-MWCNTs-β-CD composite possessed a loose and porous structure. The AFM analyses were consistent with the SEM results.

Fig. 3
figure 3

AFM images of the electrodes. a Bare GCE. b PtNPs-GNPs-MWCNTs-β-CD/GCE

Full size image

FT-IR spectroscopy

GNPs, MWCNTs, β-CD, and PtNPs-GNPs-MWCNTs-β-CD were characterized by FT-IR spectroscopy (Fig. 4). GNPs, MWCNTs, and β-CD exhibited characteristic absorption peaks at ca. 3421 cm−1, which correspond to the stretching vibration of O–H bonds, and ca. 1075 cm−1, which could be attributed to the asymmetric stretching vibration of C–O bonds. In addition, β-CD showed typical absorption peaks at 2910 cm−1, 1395 cm−1, and 1640 cm−1 which could be ascribed to the stretching vibration and bending vibration of C–H bonds, and the bending vibration of O–H bonds, respectively. After assembly, the PtNPs-GNPs-MWCNTs-β-CD composite also showed the corresponding absorption peaks as those of GNPs, β-CD, and MWCNTs with slight wavenumber shifts, indicating possible intermolecular interactions among the components [39].

Fig. 4
figure 4

FT-IR spectra of A the carbon-based materials a GNPs, b MWCNTs, c PtNPs-GNPs-MWCNTs-β-CD and B β-CD

Full size image

Raman spectroscopy

In order to reveal more structural information about the composite, Raman spectra were recorded. The spectra of GNPs, MWCNTs, and PtNPs-GNPs-MWCNTs-β-CD in Fig. 5 each contain two characteristic peaks: the D band at 1348 cm−1 corresponds to the vibrations of carbon atoms with dangling bonds in the graphene edge (or a graphene piece that contains defects), and the G band at 1577 cm−1 represents the vibration of sp2 hybrid carbon atoms in a two-dimensional hexagonal lattice. The intensity ratio of D band and G band (ID/IG) is usually used to evaluate the degree of disorders or defects of carbon-based materials [40]. The ID/IG values of GNPs and MWCNTs were 0.19 and 1.07, respectively, confirming the fewer defects or disorders in these two kinds of pristine carbon nanomaterials. The ID/IG value of PtNPs-GNPs-MWCNTs-β-CD is 0.27, which was between the values of GNPs and MWCNTs, indicating that the obtained PtNPs-GNPs-MWCNTs-β-CD composite possessed an integrated structure of GNPs and MWCNTs. Meanwhile, the other two components, PtNPs and β-CD, which were assembled with GNPs and MWCNTs, had a small effect on the ID/IG value [41].

Fig. 5
figure 5

Raman spectra of GNPs, MWCNTs, and PtNPs-GNPs-MWCNTs-β-CD

Full size image

Electrochemical properties of modified electrodes

Figure 6A presents cyclic voltammograms of bare GCE, PtNPs/GCE, GNPs/GCE, MWCNTs/GCE, β-CD/GCE, PtNPs-GNPs-MWCNTs/GCE, and PtNPs-GNPs-MWCNTs-β-CD/GCE in 1.0 mmol L−1 K3Fe(CN)6 in 0.1 mol L−1 KCl. The peak currents of bare GCE, PtNPs/GCE, GNPs/GCE, MWCNTs/GCE, β-CD/GCE, PtNPs-GNPs-MWCNTs/GCE, and PtNPs-GNPs-MWCNTs-β-CD/GCE gradually increased: 5.59, 7.68, 8.57, 9.37, 10.41, 12.75, and 14.25 μA, respectively. PtNPs/GCE, GNPs/GCE, and MWCNTs/GCE exhibited higher electrochemical responses for FA in comparison with bare GCE as a result of the high conductivity of GNPs, PtNPs, and MWCNTs. PtNPs-GNPs-MWCNTs/GCE exhibited higher peak current and electrochemical activity, indicating that the PtNPs-GNPs-MWCNTs composite greatly facilitated the electron transfer between the solution and the electrode owing to the synergistic effects of PtNPs, GNPs, and MWCNTs. Importantly, the peak current of PtNPs-GNPs-MWCNTs-β-CD/GCE exhibited the highest peak current and electrochemical activity, indicating that the PtNPs-GNPs-MWCNTs-β-CD composite greatly facilitated the electron transfer between the solution and the electrode owing to the synergistic effects of PtNPs, GNPs, MWCNTs, and β-CD.

Fig. 6
figure 6

A Cyclic voltammograms of different electrodes in 1.0 mmol L−1 K3[Fe(CN)6] solution (0.1 mol L−1 KCl). B Nyquist plots of different electrodes in 1.0 mmol L−1 K3[Fe(CN)6] solution (0.1 mol L−1 KCl) at open-circuit potential conditions (AC frequency range 0.1–106 Hz, AC amplitude 5.0 mV) (a bare/GCE, b PtNPs/GCE, c GNPs /GCE, d MWCNTs/GCE, e β-CD/GCE, f PtNPs-GNPs-MWCNTs/GCE, g PtNPs-GNPs-MWCNTs-β-CD/GCE). C CV of bare GCE in 1.0 mmol L−1 K3[Fe(CN)6] solution (0.1 mol L−1 KCl) at various scan rates (0.01, 0.02, 0.04, 0.08, 0.16, 0.20, 0.25, and 0.30 V s−1). D Plots of redox peak currents versus scan rates at bare GCE (Ipa oxidation peak current, Ipc reduction peak current). E CV of the PtNPs-GNPs-MWCNTs-β-CD/GCE in 1.0 mmol L−1 K3[Fe(CN)6] solution (containing 0.1 mol L−1 KCl) at various scan rates (0.01, 0.02, 0.04, 0.08, 0.16, 0.20, 0.25, and 0.3 V s−1). F Plots of the PtNPs-GNPs-MWCNTs-β-CD/GCE redox peak currents as a function of scan rates. Dependence of the redox peak potentials of bare GCE (G) and PtNPs-GNPs-MWCNTs-β-CD/GCE (H) on log ν (Epa oxidation peak potential, Epc reduction peak potential)

Full size image

The electrochemical properties of bare GCE and the modified electrodes were characterized by electrochemical impedance spectroscopy (EIS). As we can see in Fig. 6B, the charge-transfer resistance (Rct) at bare GCE (160.0 Ω) was higher than those at GNPs/GCE (154.9 Ω), PtNPs/GCE (130.0 Ω), β-CD/GCE (118.7 Ω), MWCNTs/GCE (111.3 Ω), PtNPs-GNPs-MWCNTs/GCE (110.8 Ω), and PtNPs-GNPs-MWCNTs-β-CD/GCE (105.7 Ω). Obviously, the Rct of PtNPs-GNPs-MWCNTs-β-CD/GCE was the lowest, which could be attributed to the excellent electrocatalytic activity, high conductivity, and large specific surface area of the electrode modifier, i.e., the assembled PtNPs-GNPs-MWCNTs-β-CD composite. The results were consistent with CVs.

In order to understand the roles of surface area in the electrochemical behavior, CV of bare GCE (Fig. 6C) and PtNPs-GNPs-MWCNTs-β-CD/GCE (Fig. 6E) was studied for 1.0 mmol L−1 K3[Fe(CN)6] which was in 0.1 mol L−1 KCl at various scan rates (0.01–0.30 V s−1). As shown in Fig. 6D, F, the redox peak currents (Ipa) of bare GCE and PtNPs-GNPs-MWCNTs-β-CD/GCE were linearly related to the square root of the scan rate, indicating that the mass transfer onto the electrode surface could be a diffusion-controlled process [42, 43]. The electroactive area of bare GCE and PtNPs-GNPs-MWCNTs-β-CD/GCE can be estimated using the Randles–Sevcik equation: Ipa (μA) = (2.69 × 105) n3/2 D1/2 C A ν1/2 (mV s−1)1/2 (where n is the number of transferred electrons in the redox reaction of K3[Fe(CN)6] solution of 1 mmol L−1; D is the diffusion coefficient of 6.70 × 10−6 cm2 s−1; C is the concentration of K3[Fe(CN)6] of 1 mmol L−1; A is the electroactive area (cm2), and ν is the scan rate (mV s−1)) [44]. From Fig. 6D, F, the linear equations of Ip and ν1/2 could be expressed as Ipa = 0.95 ν1/2 + 0.92 and Ipa = 2.32 ν1/2 – 2.28 at bare GCE and PtNPs-GNPs-MWCNTs-β-CD/GCE, respectively. According to the Randles–Sevcik equation, the electroactive area of bare GCE (0.0436 cm2) was lower than that of PtNPs-GNPs-MWCNTs-β-CD/GCE (0.105 cm2), indicating that electron transfer rate could be accelerated at PtNPs-GNPs-MWCNTs-β-CD/GCE.

The effects of scan rate on the peak potential were also investigated. As can be seen from Fig. 6C, E, the separations of the cathodic and anodic peaks at bare GCE and PtNPs-GNPs-MWCNTs-β-CD/GCE increased with an increase in scan rate (10 to 300 mV s−1), suggesting that the redox process was quasi-reversible [45]. The redox peak potentials of bare GCE and PtNPs-GNPs-MWCNTs-β-CD/GCE were plotted against the logarithmic function of scan rates (log ν), and two linear variations can be observed (Fig. 6G, H). According to Laviron’s theory [46], the cathodic and anodic slopes can be presented as −2.303RT/αnF and 2.303RT/(1 − α)nF, respectively [47], where F is the Faraday constant (96,485 C mol−1), n is the number of participating electrons (n = 1), α is the electron transfer coefficient, T is the temperature (298 K), and R is the ideal gas constant (8.314 J K−1 mol−1). On the basis of the slopes of the two straight lines and the equation of ka/kb = (α − 1)/α (where ka and kb are the slopes of the straight lines for Epa vs. log v and Epc vs. log v, respectively), the values of α at bare GCE and PtNPs-GNPs-MWCNTs-β-CD/GCE were calculated to be 0.50 and 0.46, respectively. Furthermore, when the condition of nΔEp < 200 mV is fulfilled, the electron transfer rate constant (ks) under a surface-controlled process can be obtained using Laviron’s equation: log ks = αlog (1 − α) + (1 − α)log α − log (RT/nFν) − nFα(1 − α)ΔEp/2.3RT (where the scan rate (v) is 200 mV s−1, ΔEp is the peak-to-peak potential separation, other parameters are the same as mentioned above) [46]. The ks of PtNPs-GNPs-MWCNTs-β-CD/GCE (0.903 s−1) was estimated to be higher than that of bare GCE (0.796 s−1), indicating that the PtNPs-GNPs-MWCNTs-β-CD composite contributed to promote the electron transfer between solution and electrode [48].

In addition, the heterogeneous electron transmission rate (k0) of a quasi-reversible system was estimated by the Nicholson method [49, 50] by using the following equation: k0 = ψ [(πDnνF/(RT)]1/2 (where ψ is the dimensionless rate parameter of CV related to ΔEp: ψ = (−0.628 + 0.0021ΔEp)/(1 − 0.017ΔEp), other parameters are the same as mentioned above). The values of k0 at different scan rates calculated by substituting each parameter are listed in Table 1. The average k0 of PtNPs-GNPs-MWCNTs-β-CD/GCE (3.10 × 10−3 cm s−1) was higher than that of bare GCE (2.13 × 10−3 cm s−1), indicating that PtNPs-GNPs-MWCNTs-β-CD possessed higher electrochemical reactivity [50].

Table 1 Bare GCE and PtNPs-GNPs-MWCNTs-β-CD/GCE of ΔEp and ψ in the Nernst system at 25 °C and the calculated k0
Full size table

Electrochemical behavior of FA

The electrochemical detection of FA at PtNPs-GNPs-MWCNTs-β-CD/GCE by different electrochemical methods including DPV, SWV, and CV was investigated for comparison (Fig. 7A). The peak current shape of FA by CV was much better than those in DPV and SWV. Moreover, the utilization of DPV and SWV methods would cause the desquamation of coated PtNPs-GNPs-MWCNTs-β-CD film, resulting in severe instability of the electroanalytical data. Therefore, the CV method was employed in the following electrochemical experiments.

Fig. 7
figure 7

A Electrochemical behavior of 0.1 mmol L−1 FA (0.1 M PBS, pH 7.0) using different electrochemical techniques. B CVs of different electrodes in 0.10 mmol L−1 of FA solution (0.1 mol L−1 of PBS as supporting electrolyte): a bare/GCE, b PtNPs/GCE, c GNPs/GCE d MWCNTs/GCE, e β-CD/GCE f PtNPs-GNPs/GCE, g PtNPs-MWCNTs/GCE, h PtNPs-β-CD/GCE, i GNPs-MWCNTs/GCE, j GNPs-β-CD/GCE, k MWCNTs-β-CD/GCE, l PtNPs-GNPs-MWCNTs/GCE, and m PtNPs-GNPs-MWCNTs-β-CD/GCE

Full size image

The electrochemical behavior of FA at bare GCE and the modified electrodes was investigated by CV for 0.10 mmol L−1 FA which was in 0.1 mol L−1 of PBS (pH 7.0) in the potential range of 0.2–1.0 V. As shown in Fig. 7B, oxidation of FA at the electrodes was detected and related to the formation of a double bond between C9 and N10 (Scheme 2) [51, 52]. The peak currents of 0.1 mmol L−1 FA (0.1 mol L−1 PBS, pH 7.0) at different electrodes of bare GCE, PtNPs /GCE, GNPs/GCE MWCNTs/GCE, β-CD/GCE, PtNPs-GNPs/GCE, PtNPs-MWCNTs/GCE, PtNPs-β-CD/GCE, GNPs-MWCNTs/GCE, GNPs-β-CD/GCE, MWCNTs-β-CD/GCE, PtNPs-GNPs-MWCNTs/GCE, and PtNPs-GNPs-MWCNTs-β-CD/GCE were 0.78, 1.58, 1.91, 1.57, 0.99, 2.20, 1.86, 1.14, 2.78, 2.16, 4.49, and 4.65 μA, respectively. The peak currents at the modified electrodes were greatly enhanced owing to the larger specific surface area and the higher conductivity of the working electrodes compared to bare GCE [53]. The aforementioned modified electrode also showed higher oxidation current responses for FA in comparison with bare GCE owing to the high conductivity of GNPs, PtNPs, and MWCNTs. In addition, the peak current at β-CD/GCE was lower, indicating that the insulating β-CD film would prevent the electron transfer rates. Importantly, when β-CD is present on the surface of GCE, FA would accumulate (preconcentrated) in the β-CD molecular cavities, so the surface concentration of FA is higher than that in the bulk solution. Therefore, the CV peak currents in the presence of β-CD on the surface of GCE are higher than those in the absence of β-CD. Moreover, the highest peak current could be obtained at PtNPs-GNPs-MWCNTs-β-CD/GCE, indicating that it possessed excellent electrocatalytic activity towards FA. The high conductivity and the large electroactive surface area of PtNPs-GNPs-MWCNTs-β-CD/GCE, as well as the synergistic effects of PtNPs, GNPs, MWCNTs, and β-CD were beneficial to the electrochemical detection of FA [54].

Scheme 2
scheme 2

Structure of FA

Full size image

Optimization of analytical conditions

Effects of amount of coated PtNPs-GNPs-MWCNTs-β-CD

The electrochemical responses of FA at modified GCE with various amounts of PtNPs-GNPs-MWCNTs-β-CD were investigated by CV. As we can see in Fig. 8, the peak current of FA gradually increased to a maximum value and then decreased when the amount of coated PtNPs-GNPs-MWCNTs-β-CD gradually from 5.0 μL to 10.0 μL. The electrode cannot be completely covered if the amount of coated PtNPs-GNPs-MWCNTs-β-CD is too small, and the peak current response would decrease. An excess amount of coated PtNPs-GNPs-MWCNTs-β-CD would increase the thickness of the film, and the peak current response at PtNPs-GNPs-MWCNTs-β-CD/GCE decreased because of the hindered electron transfer between the electrode and FA. The peak current of FA at the modified GCE reached to a maximum when the amount of coated PtNPs-GNPs-MWCNTs-β-CD composite was 7.0 μL. Thus, the optimal amount of coated PtNPs-GNPs-MWCNTs-β-CD suspension was set as 7.0 μL.

Fig. 8
figure 8

Effects of amount of coated PtNPs-GNPs-MWCNTs-β-CD on the electrochemical responses of modified GCE toward FA

Full size image

Effects of pH

The pH of buffer solution (0.1 mol L−1 of PBS) has an important impact on the electrochemical behavior of FA at the modified GCE. The electrochemical detection of FA at PtNPs-GNPs-MWCNTs-β-CD/GCE in PBS solutions with different pH values of 4.0, 5.0, 6.0, 7.0, 8.0, and 9.0 was investigated (Fig. 9). The oxidation peak currents of FA gradually increased with an increase in pH value of PBS and reached a maximum at pH 7.0 and then declined thereafter. FA was unionized in an acid medium and not susceptible to electrochemical oxidation because of its low pKa1 (2.3) [55]. The largest oxidation peak current could be obtained at PtNPs-GNPs-MWCNTs-β-CD/GCE for FA at pH 7.0, indicating that FA underwent an efficient electrochemical oxidation. Anionic forms of FA would appear when the pH is higher than 7.0 (pKa2 = 8.3), again resulting in a decrease in oxidation peak current [56]. Therefore, PBS of pH 7.0 was used in the following experiments.

Fig. 9
figure 9

a CV behavior of 0.1 mmol L−1 FA (containing 0.1 mol L−1 of PBS) with different pH values. b Effects of pH value (4.0, 5.0, 6.0, 7.0, 8.0, and 9.0) on peak current response of 0.1 mmol L−1 FA (containing 0.1 mol L−1 of PBS)

Full size image

Effects of scan rate

To evaluate the effect of electrochemical surface properties of PtNPs-GNPs-MWCNTs-β-CD/GCE on determination of FA, CV investigation was performed at various scan rates (0.01–0.3 V s−1) in 0.1 mmol L−1 FA in 0.1 mol L−1 PBS (pH 7.0) at a potential range of 0.2–1.0 V. As shown in Fig. 10, the oxidation peak current of FA increased with an increasing scan rate, and the linear regression equation could be expressed as Ipa = 19.18 ν1/2 − 0.33 (R2 = 0.9906). The oxidation peak current was proportional to the square root of scan rate (ν1/2), and the current at the modified electrodes conformed to the Randles–Sevcik equation [57]. The results showed that the reversible redox of FA at PtNPs-GNPs-MWCNTs-β-CD/GCE might be controlled by the rate of bulk diffusion [58].

Fig. 10
figure 10

a Cyclic voltammograms of PtNPs-GNPs-MWCNTs-β-CD/GCE for FA (0.1 mmol L−1) solution at different scan rates (0.01, 0.02, 0.05, 0.08, 0.10, 0.12, 0.20, 0.25, and 0.30 V s−1). b Plots of oxidation peak current versus square root of scan rate (ν1/2) at PtNPs-GNPs-MWCNTs-β-CD/GCE

Full size image

Determination of FA

Under the optimized experimental conditions, a series of FA solutions with different concentrations (0.02–0.50 mmol L−1) were tested by CV at PtNPs-GNPs-MWCNTs-β-CD/GCE in 0.1 mol L−1 of PBS (pH 7.0) (Fig. 11). The peak currents linearly increased with an increase in FA concentrations in the range 0.02–0.50 mmol L−1 (Fig. 11, inset). A linear equation of Ipa (μA) = 24.83 CFA (mmol L−1) + 0.097 (R2 = 0.9942) could be obtained, and the limit of detection (LOD) was estimated to be 0.48 μmol L−1 (signal-to-noise ratio = 3).

Fig. 11
figure 11

CV behavior of PtNPs-GNPs-MWCNTs-β-CD/GCE in FA (0.1 mol L−1 PBS pH 7.0) in the concentration range 0.02–0.50 mmol L−1 (inset plot of peak current versus FA concentration)

Full size image

The LOD and the linear dynamic concentration range of PtNPs-GNPs-MWCNTs-β-CD/GCE for the detection of FA were compared with some previously reported electrodes. As shown in Table 2, PtNPs-GNPs-MWCNTs-β-CD/GCE possessed lower LOD and wider linearity range. Therefore, the developed electrode is of considerable practical value for measuring FA owing to its advantages including greater simplicity, higher sensitivity, and wider linear range.

Table 2 Comparison of previously reported electrochemical sensors and PtNPs-GNPs-MWCNTs-β-CD/GCE for detection of FA
Full size table

Anti-interference ability and stability

The anti-interference ability of PtNPs-GNPs-MWCNTs-β-CD/GCE for detection of FA was evaluated. The influence of some potential coexisting interfering species including NaCl (5.0 mmol L−1), KCl (5.0 mmol L−1), MgCl2 (5.0 mmol L−1), CaCl2 (5.0 mmol L−1), Zn(NO3)2 (5.0 mmol L−1), glucose (1.0 mmol L−1), and glutamate (1.0 mmol L−1) with FA (0.1 mmol L−1) in 0.1 mol L−1 PBS at PtNPs-GNPs-MWCNTs-β-CD/GCE was investigated. Generally, the current ratio (%) is used to evaluate the anti-interference ability of the electrode, and the current ratio (%) is defined as the difference between the currents of the target object measured before and after the addition of the interfering substance. Figure 12 shows that these potential interfering agents at 10-fold or 50-fold higher concentrations than that of FA had little influence on the response signals of FA and the current ratios (%) were less than 6.16%. The results indicated that the proposed PtNPs-GNPs-MWCNTs-β-CD/GCE possessed excellent anti-interference ability.

Fig. 12
figure 12

Amperometric current response of PtNPs-GNPs-MWCNTs-β-CD/GCE for 0.1 mM FA containing some potential interfering substances including NaCl (5.0 mmol L−1), KCl (5.0 mmol L−1), MgCl2 (5.0 mmol L−1), CaCl2 (5.0 mmol L−1), Zn(NO3)2 (5.0 mmol L−1), glucose (1.0 mmol L−1), and glutamate (1.0 mmol L−1) in PBS (0.1 mol L−1, pH 7.0)

Full size image

The stability of PtNPs-GNPs-MWCNTs-β-CD/GCE was studied by CV in 0.1 mmol L−1 FA. The intra-day and inter-day precisions of PtNPs-GNPs-MWCNTs-β-CD/GCE for FA were investigated for 5 times in one day and in five consecutive days at the same electrode, respectively. Then the relative standard deviations (RSD) of intra-day precision and inter-day precision (n = 5) of PtNPs-GNPs-MWCNTs-β-CD/GCE toward FA were 5.19% and 6.07%, respectively. Therefore, the developed PtNPs-GNPs-MWCNTs-β-CD/GCE exhibited good stability and reproducibility for the determination of FA.

Determination of FA in real samples

In order to evaluate the sensitivity and practical applicability of the proposed PtNPs-GNPs-MWCNTs-β-CD/GCE, the determination of FA in actual samples, FA tablets, was performed. FA tablets (0.40 mg per pill) were purchased from a local pharmacy (Changsha, China). FA tablets were crushed and dissolved in 10 mL of 0.1 mol L−1 PBS (pH 7.0), and the standard addition method was used for the accurate determination. The spiked recoveries were in the range 97.55–102.96% and RSDs were in the range 2.07–4.64% (Table 3). Therefore, the proposed PtNPs-GNPs-MWCNTs-β-CD/GCE can be used for the determination of FA in real samples with relatively high accuracy and precision and relatively high linear correlation.

Table 3 Determination of FA in FA tablet (n = 3)
Full size table

Conclusion

In this work, a novel nanocomposite-based sensor, PtNPs-GNPs-MWCNTs-β-CD/GCE, was successfully fabricated and used for the electrochemical detection of FA by CV. Under the optimum experimental conditions, PtNPs-GNPs-MWCNTs-β-CD/GCE showed higher electrochemical responses for FA in 0.1 mol L−1 PBS solution, and relatively high linear correlation between oxidation peak current and FA concentration in the range of 0.02–0.50 mmol L−1 could be obtained. The LOD was evaluated to be 0.48 μmol L−1. The recoveries of determination of FA in FA tablets were calculated to be 97.55–102.96%. Therefore, PtNPs-GNPs-MWCNTs-β-CD/GCE could be used to accurately determine FA, and it provides a platform for the detection of FA in real samples.