Introduction

Conductive polymers have attracted a lot of attention due to their numerous potential applications in electronic devices and chemical, biomedical, and biosensing materials [1, 2]. Polyaniline (PANI), as one of the most important conductive polymers, has been extensively investigated because of its good electronic and optical properties, ease of synthesis, and good environmental stability [3–5]. In the case of sensor applications, nanostructured PANI has great sensitivity and fast response time due to its highly effective surface area and short penetration depth for target molecules [6]. However, the pure PANI is electroactive only when the pH value is under 4.0 [7, 8]. This is an obstacle to its application in sensor which normally requires a neutral environment.

In recent years, the introduction of inorganic nanoparticles (NPs) into conducting polymers has attracted attention in material science since both are simple and easy to fabricate [9]. Particularly, nanosized colloidal Au particles are considered to be excellent materials in biosensing applications due to their large specific surface area, good biocompatibility, excellent conducting properties, great improvement of the electron transfer rate, and enhancement of the capability for surface immobilization [10, 11]. The combination of PANI with Au NPs will bring enhanced performance, such as larger surface area, higher conductivity, and better biocompatibility [12–14].

Nicotinamide adenine dinucleotide (NADH) is an important coenzyme in the metabolic processes, which plays a significant role as a cofactor for more than 300 dehydrogenase enzymes in the electron transfer chain in every cell [15]. As the NAD+/NADH redox couple is involved in several hundreds of enzymatic reactions of the biological system, a highly sensitive and selective sensor is of great importance for the electrochemical detection of NADH [16]. The reversible potential of the NAD+/NADH couple is estimated to be −0.56 V [vs. saturated calomel electrode (SCE) at 25 °C], while the formal potential of the couple in neutral pH has a significant overpotential as large as 1.0 V for the direct electrochemical oxidation at bare electrodes [17, 18]. Thus, the design of the sensor for the detection of NADH at low potential in the neutral environment is of immense practical interest and has attracted great attention from the chemical community.

In this work, a tube-like PANI/Au nanocomposite was synthesized. Combining the advantages of PANI nanotubes and Au NPs, the obtained nanocomposite exhibited good electroactivity, satisfactory biocompatibility, and high electrical conductivity. Herein, the PANI/Au nanotube was modified on the surface of a glassy carbon electrode (GCE) to construct a sensor which could detect NADH sensitively at low potential of around 0.52 V in pH 7.0 phosphate buffer solution (PBS). The fabricated sensor showed good repeatability, stability, selectivity, and high catalytic activity for the electrochemical oxidation of NADH.

Experimental

Materials

Aniline, hydrochloric acid (HCl), ammonium persulfate [(NH4)2S2O8, APS], manganese sulfate monohydrate (MnSO4·H2O), potassium borohydride (KHB4), trisodium citrate, glucose (GC), and HAuC14 were purchased from the Shanghai Chemical Reagent Company. Ascorbic acid (AA) and NADH were purchased from Sigma Chemicals. All the reagents were of analytical reagent grade and used as received without further treatment. PBS was prepared from NaH2PO4 (0.1 M) and Na2HPO4 (0.1 M), and the pH value was adjusted with 0.1 M H3PO4 and NaOH solutions. Freshly prepared AA, GC, and NADH solutions were used for all experiments.

Preparation of PANI nanotubes

MnO2 nanowires were prepared according to the previously reported method [19]. For the preparation of PANI, 0.5 mM aniline monomer was added into 15 mL of 1 M HCl solution, and then, 15 mL HCl aqueous solution containing equimolar APS was slowly dropped into the above solution under stirring at room temperature. Then, 0.1 g of MnO2 was added to the above mixture under stirring. The reaction was allowed to proceed for 12 h, and the final product was centrifuged and washed with double distilled water and ethanol, and then dried overnight at 60 °C.

Preparation of PANI/Au nanocomposites

The Au colloid was prepared by reducing HAuCl4 aqueous solution with trisodium citrate and KBH4 [20]. The average diameter of the prepared Au NPs is about 5 nm. The PANI nanotubes were added to the Au colloid at a concentration of 1.0 and 2.0 mg/mL under ultrasonication for 30 min and then stirred for 1 h. Au NPs were absorbed on the surface of PANI, leading to the formation of a PANI/Au composite. By adjusting the concentration of PANI in the Au colloid, the amount of Au NPs adsorbed on the PANI surface could be controlled. And the products prepared by 1 mg PANI in 1 mL Au colloid and 2 mg PANI in 1 mL Au colloid are named as PANI/Au(1) and PANI/Au(2), respectively.

Preparation of PANI/Au- and PANI-modified glassy carbon electrodes

The GCE (3 mm in diameter) was mechanically polished with 1.0, 0.3, and 0.05 μm alumina slurry on a polish cloth, rinsed with double distilled water, briefly cleaned in an ultrasonic bath, and then dried at room temperature. PANI or PANI/Au composite nanotubes were dispersed in double distilled water to form a 2.0-mg/mL solution and ultrasonically treated for 30 min. Eight microliters of the above colloidal solution was dropped onto the pretreated GCE surface and allowed to dry under ambient conditions.

Characterization

The morphologies were investigated by transmission electron microscopy (TEM, JEOLJEM-200CX). X-ray diffraction (XRD) patterns were taken on a Philips X’Pert X-ray diffractometer with a CuKα X-ray source. Fourier transform infrared (FTIR) spectroscopic measurements were performed on a Bruker model Vector 22 Fourier transform spectrometer using KBr pressed disks. Ultraviolet visible (UV-vis) absorption spectrum was recorded on a UV-2401PC spectrometer. Thermogravimetric analysis (TGA) was carried out on a Shimadzu TGA-50 instrument from room temperature to 800 °C at a heating rate of 10 °C min−1 in an air atmosphere. Electrochemical experiments were conducted with a CHI660C workstation (Shanghai Chenhua, Shanghai) in a three-electrode system. All electrochemical experiments were performed in a cell containing 20.0 mL of PBS (0.1 M) at room temperature, using a coiled platinum wire as the auxiliary electrode, a SCE as the reference electrode, and the PANI or PANI/Au-modified electrode as the working electrode.

Results and discussion

The tube-like PANI was synthesized successfully through a self-degradable template method. During the polymerization of aniline monomer, the MnO2 nanowire is degraded automatically resulting in the formation of PANI nanotubes. Figure 1a shows the TEM images of MnO2 nanowires. The obtained MnO2 nanowires have smooth surfaces with the diameter of about 70 to 120 nm. The TEM image of the prepared PANI nanotube is shown in Fig. 1b. The inner diameter of the PANI nanotubes is in agreement with the diameter of MnO2 templates. When the PANI nanotubes were introduced into the Au colloid, Au-decorated PANI composites were formed through an electrostatic assembly method. PANI is a kind of positively polycationic polymer, while the surface of citrate-stabilized Au NPs was electronegative. As a result, they can attract each other via an electrostatic effect. The TEM image of PANI/Au(1) nanocomposite is shown in Fig. 1c. The tube-like structured material with a rough surface should be the PANI nanotube, and the dark spots dispersed on it uniformly are Au NPs which can be clearly seen on the edge of PANI nanotubes and they are partly marked in red in Fig. 1c. The average diameter of the prepared Au NPs is about 5 nm.

Fig. 1
figure 1

TEM images of a MnO2 nanowires, b PANI nanotubes, and c PANI/Au(1) composite nanotubes

Full size image

XRD patterns of the PANI and PANI/Au composites are displayed in Fig. 2. The peak centered at 2θ = 25° may be ascribed to the periodicity parallel to the polymer chain [21]. Four additional peaks at 38°, 44°, 65°, and 78° representing Bragg reflections from (111), (200), (220), and (311) planes of Au, respectively, are observed showing the existence of Au NPs in the composites [22]. Besides, Fig. 2 also suggests that the relative diffraction peak intensity for Au NPs is increased as the concentration of PANI nanotubes in Au colloid decreased. The diffraction peak intensity for Au NPs in Fig. 2a is increased compared with that in Fig. 2b. When the concentration of PANI tubes in the Au colloid is decreased, the content of Au NPs in the composite would be increased, leading to an increase in the relative peak intensity.

Fig. 2
figure 2

XRD patterns of a PANI/Au(1) composite nanotubes, b PANI/Au(2) composite nanotubes, and c PANI nanotubes

Full size image

The nature of the samples was further studied by FTIR and UV-vis. Figure 3a represents the FTIR spectra of pure PANI nanotubes and PANI/Au nanocomposites with different contents of Au NPs. It shows that the FTIR spectra of composites are identical to that of pure PANI. In general, the two bands appearing at 1,582 and 1,494 cm−1 are associated to the stretching vibration of quinoid and benzenoid rings, respectively [23]. The peaks at 1,303 and 1,246 cm−1 are related to the C–N and C═N stretching modes, respectively [24, 25]. The peak at 1,139 cm−1 is assigned to the in-plane bending of C–H [20], and the peak at 827 cm−1 corresponds to an aromatic C–H out-of-plane bending vibration [26].

Fig. 3
figure 3

FTIR spectra (a) and UV-vis absorption spectra (b) of a PANI/Au(1) composite nanotubes, b PANI/Au(2) composite nanotubes, and c PANI nanotubes

Full size image

A typical UV-vis absorption spectrum of PANI is shown in Fig. 3b(c). The peaks correspond to electronic transitions from valence band to polaron band characteristic of the PANI nanotubes at around 363, 438, and 878 nm (the broader one), which can be attributed to the benzenoid π-π* electronic transition, p-π* electronic transition, and the π to the localized polaron band, respectively [23, 26–29]. The characteristic peak of the citrate-stabilized colloidal Au NPs should appear at 510 nm and is caused by the surface plasmon resonance [25]. However, as it is shown in Fig. 3b, the surface plasmon resonance peak of Au NPs is not observed after Au NPs were adsorbed on the PANI surface [Fig. 3b(a, b)]. The loss of the surface plasmon resonance peak at 510 nm may be caused by the change of UV-vis absorption spectroscopic measurement environment since the surface plasmon resonance bands of metal NPs are sensitive to their surrounding environment [25, 30, 31]. Additionally, the strong absorption of conducting polymer can also lead to the lack of observation of a peak in the same region. As a result, the reason that the plasmon resonance band of Au NPs is not visible might be due to the overlap of absorption intensity by the surrounding PANI.

Figure 4 presents the TGA curves of the PANI/Au composites and intrinsic PANI. The curves a and b show the weight loss of PANI/Au composites with different contents of Au NPs. Three-stage weight losses are observed in both PANI/Au samples. The first step from room temperature to around 100 °C represents the loss of the moisture, and the second step up to around 230 °C is attributed to the elimination of acid dopant [32, 33]. There is a continuous weight loss in the range of 300 to 620 °C which is believed to be due to the oxidation of the skeletal PANI chain structure [24, 32]. In the case of intrinsic PANI, curve c shows a 17.5 % weight loss in the region of room temperature to around 380 °C which is assigned to the loss of the water and dopants. And another weight loss in the region of 400 to 650 °C is the degradation of the PANI chain. The total weight loss of intrinsic PANI is 100 % in the experimental conditions. Under the experiment conditions, there is no weight loss of Au NPs. The residual weight (in percent), which refers to the weight percentage content of Au NPs in the composites, is 13.5 and 6.3 % for PANI/Au(1) and PANI/Au(2), respectively.

Fig. 4
figure 4

TGA spectra of a PANI/Au(1) composite nanotubes, b PANI/Au(2) composite nanotubes, and c PANI nanotubes

Full size image

The electrochemical activity of the PANI/Au nanotube was detected by cyclic voltammogram (CV) experiments. Figure 5 depicts the CVs of PANI/Au(1) nanocomposite-modified GCE in buffer solutions with different pH values ranging from 1.0 to 7.0. It is acknowledged that PANI exists in three well-defined oxidation states: leucoemeraldine, emeraldine, and pernigraniline. As shown in Fig. 5, two separate redox peaks are obtained in pH 1 and 2 PBS. The first and second oxidation waves correspond to the transition of leucoemeraldine to emeraldine salt and emeraldine salt to pernigraniline state in acid media [20], respectively. In the leucoemeraldine state, all the nitrogen atoms are amines, but in pernigraniline, the nitrogen atoms are imines. In the different states of PANI, the amine/imine ratio is different. Furthermore, emeraldine can be in its base or salt form depending on the pH value. With the increase of pH value of the solution, however, the two pairs of peaks move closer and finally become only one pair of broad redox peaks when the pH value is 4. There is still a pair of peaks when the pH values are 5 and 6. This redox peak is the overlap of the two redox processes of PANI in acidic conditions [34].

Fig. 5
figure 5

Cyclic voltammograms of PANI/Au(1) composite nanotube-modified GCE in different pH values of 0.1 M PBS at the scan rate of 100 mV/s

Full size image

As the pH value increased, the peak current decreased and the separation between the anodic and cathodic peaks increased, indicating reduction of the electroactivity of PANI. As we know, pure PANI is electroactive only when the pH is under 4.0 [7, 8]. This greatly restricts its applicability in bioelectrochemisty which normally requires a neutral environment. In this experiment, however, the PANI decorated with Au NPs brings a great improvement to its electroactive ability. As shown in Fig. 5, the PANI/Au composite-modified GCE shows redox peaks both in acidic solutions and a near neutral environment (pH 6.0). The electroactivity of PANI can be improved through doping with negatively charged groups [35], and the Au NPs with sodium citrate can provide the negative charges. In addition, the different synthetic method may cause different electroactive abilities of PANI. In this work, the high doping degree of chloride ions and the perfect tube-like structure may be another reason for the good electroactivity of the PANI nanotubes when pH is higher than 4.

Based on their good biocompatibility, high conductivity, and excellent electrochemical behavior, PANI/Au nanocomposites were immobilized onto the surface of a GCE to be developed as a sensor and applied for the detection of NADH. The electrocatalytic response to the oxidation of NADH was detected via cyclic CVs and differential pulse voltammograms (DPV). The GCE was previously tested in pH 7.0 PBS buffer before the PANI/Au nanotubes were drop-coated on it. The PANI/Au nanocomposite-modified GCE was immersed in the electrolyte solution for 30 min prior to the measurement to ensure the diffusion of the solution to the interlayer space and permit better ionic exchange.

Figure 6 displays the electrocatalytic response of the PANI/Au(1), PANI/Au(2), and PANI-modified GCE and bare GCE in 0.1 M PBS of pH 7.0 with and without the addition of 1.0 mM NADH. As it could be seen in Fig. 6, there was no obvious current in the blank PBS without the addition of NADH. Nevertheless, when the NADH was added, all the modified GCE as well as the bare one presented an oxidation current which increased in the order of bare GCE (Fig. 6d), PANI-modified GCE (Fig. 6c), PANI/Au(2)-modified GCE (Fig. 6b), and PANI/Au(1)-modified GCE (Fig. 6a). Thus, it was confirmed that the Au NPs immobilized on the PANI film caused the increase of catalytic current, which indicated that the larger the amount of Au NPs absorbed on the PANI nanotubes, the higher the catalytic current that would be obtained. These results implied that the Au NPs could lead to the acceleration of electron transfer rate and enhance the electrocatalytic activity of the modified electrode. Here, Au NPs, acting as nanoscale electrodes, promoted the electron transfer rate and provided the necessary pathways between the analyte and electrode [36]. The PANI matrix served mainly as points of attachment for Au NPs and an electron mediator, shuttling electrons between Au NPs and the electrode surface.

Fig. 6
figure 6

Cyclic voltammograms of different materials-modified electrodes in 0.1 M PBS (pH 7.0) containing 1 mM NADH and 0 mM NADH at the scan rate of 100 mV/s, respectively. a PANI/Au(1) composite nanotubes, b PANI/Au(2) composite nanotubes, c PANI nanotube-modified GCE, and d bare GCE

Full size image

The interference from ascorbic acid (AA) and glucose (GC) is an obstacle to the detection of NADH as they have a similar oxidation potential [37], which leads to an overlapping voltammetric response. To test the sensitivity and selectivity of the prepared sensor in the presence of this interference, simultaneous determination of NADH, AA, and GC at PANI/Au(1)-modified GCE was carried out using DPV. As shown in Fig. 7, a large oxidation peak at 0.52 V is obtained in curve a corresponding to the addition of NADH. However, there is no obvious oxidation peak after the addition of AA and GC, indicating that NADH can be detected in the presence of AA or GC and the anti-interference ability of the sensor is enhanced.

Fig. 7
figure 7

Differential pulse voltammograms of PANI/Au(1) composite nanotube-modified GCE in 0.1 M PBS (pH 7.0) containing a 1 mM NADH, b 1 mM AA, c 1 mM glucose, and d blank at the scan rate of 100 mV/s

Full size image

The DPV of the PANI/Au(1)-modified GCE in PBS pH 7.0 with different concentrations of NADH is shown in Fig. 8. Different concentrations of NADH solution were added to the buffer solution successively. After the mixed NADH was dispersed, CV was performed until the currents showed no change, and then, DPV experiments were carried out. It could be observed that with the addition of the NADH, a peak at around 0.53 V emerged. The peak current gradually increased and was proportional to the NADH concentrations as shown in the inset in Fig. 8. The results showed that the PANI/Au(1)-based sensor displays a wide linear range from 4 × 10−4 to 8 × 10−3 M with a correlation coefficient of 0.9946, which could be considered as largely extended comparing with the literatures [38–41]. A low detection limit of 0.5 × 10−7 M was obtained at a signal to noise ratio of 3 which was lower than that of the reported literatures [41–43]. The high performance in the detection of NADH indicated that the PANI/Au nanocomposites greatly enhanced efficiency of the electron transfer between NADH and the electrode [44].

Fig. 8
figure 8

Differential pulse voltammograms of PANI/Au(1) composite nanotube-modified GCE in 0.1 M PBS (pH 7.0) containing 0.30, 0.40, 0.48, 0.50, 0.56, 0.70, 0.78, 0.90, 1.00, 1.60, 2.00, 3.00, 3.60, 5.00, 6.00, 7.00, 8.00, 9.00, 9.60, and 10.0 mM NADH from bottom to top. Insert shows the relationship between the peak current and the concentration of NADH

Full size image

The repeatability and stability of the sensor were studied by successively testing 1 mM NADH for six times, the relative standard deviation (RSD) was 0.4 %, demonstrating a good repeatability. After 100 successive scanning, the response current still retained 95 % value of the initial, suggesting the good durability of this sensor. Additionally, the RSD of current signals for detection of 1 mM NADH at five independently prepared sensors was 3.8 %, showing excellent repeatability. When PANI/Au-modified GCE was stored at 4 °C, no obvious decrease in the response to NADH was observed after 7 days.

Conclusions

In this work, a PANI/Au composite with a tube-like shape was synthesized. By decorating the Au NPs onto the surface of PANI, the composite was obtained. The composite-modified electrode showed high electrocatalytic activity towards the oxidation of NADH in a neutral environment compared with that of the pure PANI-modified electrode. The fabricated NADH sensor had excellent anti-interference ability and displayed a wide linear range from 4 × 10−4 to 8 × 10−3 M with a detection limit of 0.5 × 10−7 M. Furthermore, the sensor showed good repeatability and stability.